liurui/PriceForecast
Template
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

原油月度数据新增中位数，最大值，最小值

Browse Source
This commit is contained in:
workpc 2025-10-21 10:36:16 +08:00
parent 58b4956db2
commit 2e83105d7d
27 changed files with 3572 additions and 1819 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

BIN
aisenzhecode/沥青/日度价格预测_最佳模型.pkl
View File

Binary file not shown.

862
aisenzhecode/沥青/沥青定量每日执行.py
View File

File diff suppressed because it is too large Load Diff

BIN
aisenzhecode/沥青/沥青定量预测2025年10月9日.zip Normal file
View File

Binary file not shown.

BIN
aisenzhecode/沥青/沥青数据项 - 副本.xlsx Normal file
View File

Binary file not shown.

BIN
aisenzhecode/沥青/沥青数据项.xlsx
View File

Binary file not shown.

BIN
aisenzhecode/沥青/沥青数据项2025年9月26日备份.xlsx Normal file
View File

Binary file not shown.

61
config_jingbo.py
View File

@ -93,7 +93,7 @@ data = {
ClassifyId = 1214
# # 变量定义--线上环境
# 变量定义--线上环境
# server_host = '10.200.32.39'
# login_pushreport_url = "http://10.200.32.39/jingbo-api/api/server/login"
# upload_url = "http://10.200.32.39/jingbo-api/api/analysis/reportInfo/researchUploadReportSave"
@ -218,6 +218,27 @@ ClassifyId = 1214
# }
# 套期保值正式环境start
tqbz_login_url = "http://10.200.32.39/jbsh/api/server/login"
tqbz_login_data = {
"funcModule": "login",
"funcOperation": "login",
"data": {
"account": "api_szh",
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=",
"tenantHashCode":"7145255749fb7d61263081fca3cb1469",
"terminal": "API"
}
}
query_is_trading_day_url = "http://10.200.32.39/jbsh/api/dd/futuresCalendar/checkTradeDay"
is_trading_day_data = {
"funcModule": "检查是否交易日",
"funcOperation": "检查是否交易日",
"data": "20251010"
}
# 套期保值正式环境end
# # 生产环境数据库
# host = 'rm-2zehj3r1n60ttz9x5.mysql.rds.aliyuncs.com'
# port = 3306
@ -227,7 +248,7 @@ ClassifyId = 1214
# table_name = 'v_tbl_crude_oil_warning'
# # 变量定义--测试环境
# 变量定义--测试环境
server_host = '192.168.100.53' # 内网
# server_host = '183.242.74.28' # 外网
login_pushreport_url = f"http://{server_host}:8080/jingbo-dev/api/server/login"
@ -342,9 +363,35 @@ push_waring_data_value_list_data = {
get_waring_data_value_list_data = {
"data": "8", "funcModule": "商品数据同步", "funcOperation": "同步"}
# 套期保值北京环境start
tqbz_login_url = "http://192.168.101.197:8080/jbsh/api/server/login"
tqbz_login_data = {
"funcModule": "login",
"funcOperation": "login",
"data": {
"account": "apitest",
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=",
"tenantHashCode": "c5207389b0a254149fccc69c9c68397e",
"terminal": "API"
}
}
query_is_trading_day_url = "http://192.168.101.197:8080/jbsh/api/dd/futuresCalendar/checkTradeDay"
is_trading_day_data = {
"funcModule": "检查是否交易日",
"funcOperation": "检查是否交易日",
"data": "20251010"
}
# 套期保值北京环境end
# 八大维度数据项编码
bdwd_items = {
'ciri': 'yyycbdwdcr',
'cierri': 'yyycbdwdcer',
'cisanri': 'yyycbdwdcsanr',
'cisiri': 'yyycbdwdcsir',
'benzhou': 'yyycbdwdbz',
'cizhou': 'yyycbdwdcz',
'gezhou': 'yyycbdwdgz',
@ -373,15 +420,17 @@ DEFAULT_CONFIG = {
'create_user': 'admin',
'create_date': datetime.datetime.now(),
'update_user': 'admin',
'update_date': datetime.datetime.now()
'update_date': datetime.datetime.now(),
'oil_code': 'CRUDE',
'oil_name': '原油',
}
# 开关
is_train = True # 是否训练
is_train = False # 是否训练
is_debug = False # 是否调试
is_eta = True # 是否使用eta接口
is_market = True # 是否通过市场信息平台获取特征 ,在is_eta 为true 的情况下生效
is_eta = False # 是否使用eta接口
is_market = False # 是否通过市场信息平台获取特征 ,在is_eta 为true 的情况下生效
is_timefurture = True # 是否使用时间特征
is_fivemodels = False # 是否使用之前保存的最佳的5个模型
is_edbcode = False # 特征使用edbcoding列表中的

73
config_jingbo_yuedu.py
View File

@ -39,6 +39,7 @@ edbnamelist = [
# '新加坡(含硫0.05%) 柴油现货价','柴油：10ppm：国际市场：FOB中间价：新加坡（日）','Bloomberg Commodity Fair Value Singapore Mogas 92 Swap Month 1','97#汽油FOB新加坡现货价','无铅汽油：97#：国际市场：FOB中间价：新加坡（日）'
]
edbcodenamedict = {}
# eta自有数据指标编码 次月,次二月，次三月，次四月
bdwdname = [
@ -47,6 +48,13 @@ bdwdname = [
'次三月',
'次四月',
]
# 数据库预测结果表八大维度列名
price_columns = [
'day_price', 'week_price', 'second_week_price', 'next_week_price',
'next_month_price', 'next_february_price', 'next_march_price', 'next_april_price'
]
modelsindex = [
{
"NHITS": "SELF0000143",
@ -180,6 +188,7 @@ ClassifyId = 1214
# query_data_list_item_nos_url = f"http://{server_host}/jingbo-api/api/warehouse/dwDataItem/queryDataListItemNos"
# # 上传数据项值
# push_data_value_list_url = f"http://{server_host}/jingbo-api/api/dw/dataValue/pushDataValueList"
# push_png_report_url = f"http://{server_host}/jingbo-api/api/analysis/reportInfo/priceForecastImg"
# login_data = {
# "data": {
@ -253,6 +262,21 @@ ClassifyId = 1214
# }
# ]
# }
# push_png_report_data = {
# "funcModule": '聚烯烃图片报告',
# "funcOperation": '上传聚烯烃PP价格预测图片报告',
# "data": {
# "groupNo": "000211",
# "updateTime": "2024-09-06 15:01:29",
# "fileBase64": '', # 文件内容base64
# "title": '2025年8月5日日度周度预测结果',
# "billNo": '',
# "pushContent": ""
# }
# }
# # 八大维度数据项编码
# bdwd_items = {
# 'ciri': '原油大数据预测|FORECAST|PRICE|T',
@ -291,6 +315,10 @@ upload_warning_url = f"http://{server_host}/jingbo-dev/api/basicBuiness/crudeOil
query_data_list_item_nos_url = f"http://{server_host}/jingbo-dev/api/warehouse/dwDataItem/queryDataListItemNos"
# 上传数据项值
push_data_value_list_url = f"http://{server_host}/jingbo-dev/api/dw/dataValue/pushDataValueList"
# 上传图片报告
push_png_report_url = f"http://{server_host}/jingbo-dev/api/analysis/reportInfo/priceForecastImg"
# 报告审核
upload_report_audit_url = f"http://{server_host}/jingbo-dev/api/analysis/reportInfo/researchImgReportSave"
login_data = {
"data": {
@ -364,6 +392,43 @@ push_data_value_list_data = {
}
]
}
push_png_report_data = {
"funcModule": '研究报告信息',
"funcOperation": '上传原油预测报告',
"data": {
"groupNo": "000161",
"updateTime": "2024-09-06 15:01:29",
"fileBase64": '', # 文件内容base64
"title": '2025年8月5日日度周度预测结果',
"billNo": '',
"pushContent": "2025年8月5日日度周度预测结果"
}
}
upload_report_audit_data = {
"groupNo": '000197', # 用户组id
"funcModule": '研究报告信息',
"funcOperation": '原油价格预测报告审核预览',
"data": {
"ownerAccount": 'arui', # 报告所属用户账号
"reportType": 'OIL_PRICE_FORECAST', # 报告类型，固定为OIL_PRICE_FORECAST
"fileName": '2000-40-5-50--100-原油指标数据.xlsx-Brent活跃合约--2024-09-06-15-01-29-预测报告.pdf', # 文件名称
"fileBase64": '', # 文件内容base64
"imgFileBase64": '', # 图片文件内容base64
"categoryNo": 'yyjgycbg', # 研究报告分类编码
"smartBusinessClassCode": 'YCJGYCBG', # 分析报告分类编码
"reportEmployeeCode": "E40116", # 报告人
"reportDeptCode": "D0044", # 报告部门
"productGroupCode": "RAW_MATERIAL", # 商品分类
"imgGroupNo": '000197', # 推送图片接口人员用户组编码
"authGroupNo":'000197', # 权限用户组编码
}
}
# 八大维度数据项编码
bdwd_items = {
'ciri': 'yyycbdwdcr',
@ -397,12 +462,14 @@ DEFAULT_CONFIG = {
'create_user': 'admin',
'create_date': datetime.datetime.now(),
'update_user': 'admin',
'update_date': datetime.datetime.now()
'update_date': datetime.datetime.now(),
'oil_code': 'CRUDE',
'oil_name': '原油',
}
# 开关
is_train = True # 是否训练
is_train = False # 是否训练
is_debug = False # 是否调试
is_eta = False # 是否使用eta接口
is_market = False # 是否通过市场信息平台获取特征 ,在is_eta 为true 的情况下生效
@ -410,7 +477,7 @@ is_timefurture = True # 是否使用时间特征
is_fivemodels = False # 是否使用之前保存的最佳的5个模型
is_edbcode = False # 特征使用edbcoding列表中的
is_edbnamelist = False # 自定义特征，对应上面的edbnamelist
is_update_eta = True # 预测结果上传到eta
is_update_eta = False # 预测结果上传到eta
is_update_report = False # 是否上传报告
is_update_warning_data = False # 是否上传预警数据
is_update_predict_value = True # 是否上传预测值到市场信息平台

13
config_jingbo_zhoudu.py
View File

@ -306,6 +306,9 @@ push_data_value_list_data = {
# 八大维度数据项编码
bdwd_items = {
'ciri': 'yyycbdwdcr',
'cierri': 'yyycbdwdcer',
'cisanri': 'yyycbdwdcsanr',
'cisiri': 'yyycbdwdcsir',
'benzhou': 'yyycbdwdbz',
'cizhou': 'yyycbdwdcz',
'gezhou': 'yyycbdwdgz',
@ -324,7 +327,6 @@ password = '123456'
dbname = 'jingbo_test'
table_name = 'v_tbl_crude_oil_warning'
DEFAULT_CONFIG = {
'feature_factor_frequency': 'D',
'strategy_id': 1,
@ -335,12 +337,15 @@ DEFAULT_CONFIG = {
'create_user': 'admin',
'create_date': datetime.datetime.now(),
'update_user': 'admin',
'update_date': datetime.datetime.now()
'update_date': datetime.datetime.now(),
'oil_code': 'CRUDE',
'oil_name': '原油',
}
# 开关
is_train = True # 是否训练
is_train = False # 是否训练
is_debug = False # 是否调试
is_eta = False # 是否使用eta接口
is_market = False # 是否通过市场信息平台获取特征 ,在is_eta 为true 的情况下生效
@ -348,7 +353,7 @@ is_timefurture = True # 是否使用时间特征
is_fivemodels = False # 是否使用之前保存的最佳的5个模型
is_edbcode = False # 特征使用edbcoding列表中的
is_edbnamelist = False # 自定义特征，对应上面的edbnamelist
is_update_eta = True # 预测结果上传到eta
is_update_eta = False # 预测结果上传到eta
is_update_report = False # 是否上传报告
is_update_warning_data = False # 是否上传预警数据
is_update_predict_value = True # 是否上传预测值到市场信息平台

268
config_juxiting.py
View File

@ -155,7 +155,6 @@ push_waring_data_value_list_url = f"http://{server_host}/jingbo-api/api/basicBui
# 获取预警数据中取消订阅指标ID
get_waring_data_value_list_url = f"http://{server_host}/jingbo-api/api/basicBuiness/crudeOilWarning/dataList"
login_data = {
"data": {
"account": "api_dev",
@ -254,6 +253,27 @@ push_waring_data_value_list_data = {
get_waring_data_value_list_data = {
"data": "9", "funcModule": "商品数据同步", "funcOperation": "同步"}
# 套期保值正式环境start
tqbz_login_url = "http://10.200.32.39/jbsh/api/server/login"
tqbz_login_data = {
"funcModule": "login",
"funcOperation": "login",
"data": {
"account": "api_szh",
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=",
"tenantHashCode":"7145255749fb7d61263081fca3cb1469",
"terminal": "API"
}
}
query_is_trading_day_url = "http://10.200.32.39/jbsh/api/dd/futuresCalendar/checkTradeDay"
is_trading_day_data = {
"funcModule": "检查是否交易日",
"funcOperation": "检查是否交易日",
"data": "20251010"
}
# 套期保值正式环境end
# 八大维度数据项编码
bdwd_items = {
@ -281,136 +301,138 @@ bdwd_items = {
# 变量定义--测试环境
# server_host = '192.168.100.53' # 内网
# # server_host = '183.242.74.28' # 外网
# login_pushreport_url = f"http://{server_host}:8080/jingbo-dev/api/server/login"
# # 上传报告
# upload_url = f"http://{server_host}:8080/jingbo-dev/api/analysis/reportInfo/researchUploadReportSave"
# # 停更预警
# upload_warning_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/save"
# # 查询数据项编码
# query_data_list_item_nos_url = f"http://{server_host}:8080/jingbo-dev/api/warehouse/dwDataItem/queryDataListItemNos"
# # 上传数据项值
# push_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/dw/dataValue/pushDataValueList"
# # 上传停更数据到市场信息平台
# push_waring_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/crudeSaveOrupdate"
# # 获取预警数据中取消订阅指标ID
# get_waring_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/dataList"
# login_data = {
# "data": {
# "account": "api_test",
# # "password": "MmVmNzNlOWI0MmY0ZDdjZGUwNzE3ZjFiMDJiZDZjZWU=", # Shihua@123456
# "password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=", # 123456
# "tenantHashCode": "8a4577dbd919675758d57999a1e891fe",
# "terminal": "API"
# },
# "funcModule": "API",
# "funcOperation": "获取token"
# }
# upload_data = {
# "funcModule": '研究报告信息',
# "funcOperation": '上传聚烯烃PP价格预测报告',
# "data": {
# "groupNo": "000127",
# "ownerAccount": 'arui', # 报告所属用户账号
# "reportType": 'OIL_PRICE_FORECAST', # 报告类型，固定为OIL_PRICE_FORECAST
# "fileName": '2000-40-5-50--100-原油指标数据.xlsx-Brent活跃合约--2024-09-06-15-01-29-预测报告.pdf', # 文件名称
# "fileBase64": '', # 文件内容base64
# "categoryNo": 'yyjgycbg', # 研究报告分类编码
# "smartBusinessClassCode": 'JXTJGYCBG', # 分析报告分类编码
# "reportEmployeeCode": "E40116", # 报告人
# "reportDeptCode": "D0044", # 报告部门
# "productGroupCode": "RAW_MATERIAL" # 商品分类
# }
# }
# # 已弃用
# warning_data = {
# "funcModule": '原油特征停更预警',
# "funcOperation": '原油特征停更预警',
# "data": {
# "groupNo": "000127",
# 'WARNING_TYPE_NAME': '特征数据停更预警',
# 'WARNING_CONTENT': '',
# 'WARNING_DATE': ''
# }
# }
# query_data_list_item_nos_data = {
# "funcModule": "数据项",
# "funcOperation": "查询",
# "data": {
# "dateStart": "20200101",
# "dateEnd": "",
# # 数据项编码,代表 PP期货 价格
# "dataItemNoList": ["MAIN_CONFT_SETTLE_PRICE"]
# }
# }
server_host = '192.168.100.53' # 内网
# server_host = '183.242.74.28' # 外网
login_pushreport_url = f"http://{server_host}:8080/jingbo-dev/api/server/login"
# 上传报告
upload_url = f"http://{server_host}:8080/jingbo-dev/api/analysis/reportInfo/researchUploadReportSave"
# 停更预警
upload_warning_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/save"
# 查询数据项编码
query_data_list_item_nos_url = f"http://{server_host}:8080/jingbo-dev/api/warehouse/dwDataItem/queryDataListItemNos"
# 上传数据项值
push_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/dw/dataValue/pushDataValueList"
# 上传停更数据到市场信息平台
push_waring_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/crudeSaveOrupdate"
# 获取预警数据中取消订阅指标ID
get_waring_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/dataList"
# push_data_value_list_data = {
# "funcModule": "数据表信息列表",
# "funcOperation": "新增",
# "data": [
# {"dataItemNo": "91230600716676129",
# "dataDate": "20230113",
# "dataStatus": "add",
# "dataValue": 100.11
# },
# {"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
# "dataDate": "20230113",
# "dataStatus": "add",
# "dataValue": 100.55
# },
# {"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
# "dataDate": "20230113",
# "dataStatus": "add",
# "dataValue": 100.55
# }
# ]
# }
login_data = {
"data": {
"account": "api_test",
# "password": "MmVmNzNlOWI0MmY0ZDdjZGUwNzE3ZjFiMDJiZDZjZWU=", # Shihua@123456
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=", # 123456
"tenantHashCode": "8a4577dbd919675758d57999a1e891fe",
"terminal": "API"
},
"funcModule": "API",
"funcOperation": "获取token"
}
upload_data = {
"funcModule": '研究报告信息',
"funcOperation": '上传聚烯烃PP价格预测报告',
"data": {
"groupNo": "000127",
"ownerAccount": 'arui', # 报告所属用户账号
"reportType": 'OIL_PRICE_FORECAST', # 报告类型，固定为OIL_PRICE_FORECAST
"fileName": '2000-40-5-50--100-原油指标数据.xlsx-Brent活跃合约--2024-09-06-15-01-29-预测报告.pdf', # 文件名称
"fileBase64": '', # 文件内容base64
"categoryNo": 'yyjgycbg', # 研究报告分类编码
"smartBusinessClassCode": 'JXTJGYCBG', # 分析报告分类编码
"reportEmployeeCode": "E40116", # 报告人
"reportDeptCode": "D0044", # 报告部门
"productGroupCode": "RAW_MATERIAL" # 商品分类
}
}
# 已弃用
warning_data = {
"funcModule": '原油特征停更预警',
"funcOperation": '原油特征停更预警',
"data": {
"groupNo": "000127",
'WARNING_TYPE_NAME': '特征数据停更预警',
'WARNING_CONTENT': '',
'WARNING_DATE': ''
}
}
query_data_list_item_nos_data = {
"funcModule": "数据项",
"funcOperation": "查询",
"data": {
"dateStart": "20200101",
"dateEnd": "",
# 数据项编码,代表 PP期货 价格
"dataItemNoList": ["MAIN_CONFT_SETTLE_PRICE"]
}
}
# push_waring_data_value_list_data = {
# "data": {
# "crudeOilWarningDtoList": [
# {
# "lastUpdateDate": "20240501",
# "updateSuspensionCycle": 1,
# "dataSource": "9",
# "frequency": "1",
# "indicatorName": "美元指数",
# "indicatorId": "myzs001",
# "warningDate": "2024-05-13"
# }
# ],
# "dataSource": "9"
# },
# "funcModule": "商品数据同步",
# "funcOperation": "同步"
# }
push_data_value_list_data = {
"funcModule": "数据表信息列表",
"funcOperation": "新增",
"data": [
{"dataItemNo": "91230600716676129",
"dataDate": "20230113",
"dataStatus": "add",
"dataValue": 100.11
},
{"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
"dataDate": "20230113",
"dataStatus": "add",
"dataValue": 100.55
},
{"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
"dataDate": "20230113",
"dataStatus": "add",
"dataValue": 100.55
}
]
}
# get_waring_data_value_list_data = {
# "data": "9", "funcModule": "商品数据同步", "funcOperation": "同步"}
push_waring_data_value_list_data = {
"data": {
"crudeOilWarningDtoList": [
{
"lastUpdateDate": "20240501",
"updateSuspensionCycle": 1,
"dataSource": "9",
"frequency": "1",
"indicatorName": "美元指数",
"indicatorId": "myzs001",
"warningDate": "2024-05-13"
}
],
"dataSource": "9"
},
"funcModule": "商品数据同步",
"funcOperation": "同步"
}
# # 八大维度数据项编码
# bdwd_items = {
# 'ciri': 'jxtppbdwdcr',
# 'cierri': 'jxtppbdwdcer',
# 'cisanri': 'jxtppbdwdcsanr',
# 'cisiri': 'jxtppbdwdcsir',
# 'benzhou': 'jxtppbdwdbz',
# 'cizhou': 'jxtppbdwdcz',
# 'gezhou': 'jxtppbdwdgz',
# 'ciyue': 'jxtppbdwdcy',
# 'cieryue': 'jxtppbdwdcey',
# 'cisanyue': 'jxtppbdwdcsany',
# 'cisiyue': 'jxtppbdwdcsiy',
# }
get_waring_data_value_list_data = {
"data": "9", "funcModule": "商品数据同步", "funcOperation": "同步"}
# 八大维度数据项编码
bdwd_items = {
'ciri': 'jxtppbdwdcr',
'cierri': 'jxtppbdwdcer',
'cisanri': 'jxtppbdwdcsanr',
'cisiri': 'jxtppbdwdcsir',
'benzhou': 'jxtppbdwdbz',
'cizhou': 'jxtppbdwdcz',
'gezhou': 'jxtppbdwdgz',
'ciyue': 'jxtppbdwdcy',
'cieryue': 'jxtppbdwdcey',
'cisanyue': 'jxtppbdwdcsany',
'cisiyue': 'jxtppbdwdcsiy',
}
# 北京环境数据库
host = '192.168.101.27'
port = 3306

347
config_juxiting_yuedu.py
View File

@ -218,148 +218,19 @@ ClassifyId = 1161
# 变量定义--线上环境
server_host = '10.200.32.39'
login_pushreport_url = "http://10.200.32.39/jingbo-api/api/server/login"
upload_url = "http://10.200.32.39/jingbo-api/api/analysis/reportInfo/researchUploadReportSave"
upload_warning_url = "http://10.200.32.39/jingbo-api/api/basicBuiness/crudeOilWarning/save"
query_data_list_item_nos_url = f"http://{server_host}/jingbo-api/api/warehouse/dwDataItem/queryDataListItemNos"
# 上传数据项值
push_data_value_list_url = f"http://{server_host}/jingbo-api/api/dw/dataValue/pushDataValueList"
push_png_report_url = f"http://{server_host}/jingbo-api/api/analysis/reportInfo/priceForecastImg"
login_data = {
"data": {
"account": "api_dev",
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=",
"tenantHashCode": "8a4577dbd919675758d57999a1e891fe",
"terminal": "API"
},
"funcModule": "API",
"funcOperation": "获取token"
}
upload_data = {
"funcModule": '研究报告信息',
"funcOperation": '上传原油价格预测报告',
"data": {
"groupNo": '', # 用户组id
"ownerAccount": '27663', # 报告所属用户账号 27663 - 刘小朋
"reportType": 'OIL_PRICE_FORECAST', # 报告类型，固定为OIL_PRICE_FORECAST
"fileName": '', # 文件名称
"fileBase64": '', # 文件内容base64
"categoryNo": 'yyjgycbg', # 研究报告分类编码
"smartBusinessClassCode": 'YCJGYCBG', # 分析报告分类编码
"reportEmployeeCode": "E40482", # 报告人 E40482 - 管理员 0000027663 - 刘小朋
"reportDeptCode": "002000621000", # 报告部门 - 002000621000 SH期货研究部
"productGroupCode": "RAW_MATERIAL" # 商品分类
}
}
warning_data = {
"groupNo": '', # 用户组id
"funcModule": '原油特征停更预警',
"funcOperation": '原油特征停更预警',
"data": {
'WARNING_TYPE_NAME': '特征数据停更预警',
'WARNING_CONTENT': '',
'WARNING_DATE': ''
}
}
query_data_list_item_nos_data = {
"funcModule": "数据项",
"funcOperation": "查询",
"data": {
"dateStart": "20150101",
"dateEnd": "20301231",
"dataItemNoList": ["MAIN_CONFT_SETTLE_PRICE"] # 数据项编码,代表 brent最低价和最高价
}
}
push_data_value_list_data = {
"funcModule": "数据表信息列表",
"funcOperation": "新增",
"data": [
{"dataItemNo": "91230600716676129",
"dataDate": "20230113",
"dataStatus": "add",
"dataValue": 100.11
},
{"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
"dataDate": "20230113",
"dataStatus": "add",
"dataValue": 100.55
},
{"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
"dataDate": "20230113",
"dataStatus": "add",
"dataValue": 100.55
}
]
}
push_png_report_data = {
"funcModule": '聚烯烃图片报告',
"funcOperation": '上传聚烯烃PP价格预测图片报告',
"data": {
"groupNo": "000211",
"updateTime": "2024-09-06 15:01:29",
"fileBase64": '', # 文件内容base64
"title": '2025年8月5日日度周度预测结果',
"billNo": '',
}
}
# 八大维度数据项编码
bdwd_items = {
'ciri': '251889263|FORECAST|PRICE|T01',
'cierri': '251889263|FORECAST|PRICE|T02',
'cisanri': '251889263|FORECAST|PRICE|T03',
'cisiri': '251889263|FORECAST|PRICE|T04',
'benzhou': '251889263|FORECAST|PRICE|T05',
'cizhou': '251889263|FORECAST|PRICE|W_01',
'gezhou': '251889263|FORECAST|PRICE|W_02',
'ciyue': '251889263|FORECAST|PRICE|M_01',
'cieryue': '251889263|FORECAST|PRICE|M_02',
'cisanyue': '251889263|FORECAST|PRICE|M_03',
'cisiyue': '251889263|FORECAST|PRICE|M_04',
}
# 报告中八大维度数据项重命名
columnsrename = {'251889263|FORECAST|PRICE|T05': '本周', '251889263|FORECAST|PRICE|M_02': '次二月', '251889263|FORECAST|PRICE|T01': '次日', '251889263|FORECAST|PRICE|M_04': '次四月',
'251889263|FORECAST|PRICE|M_03': '次三月', '251889263|FORECAST|PRICE|M_01': '次月', '251889263|FORECAST|PRICE|W_01': '次周', '251889263|FORECAST|PRICE|W_02': '隔周', }
# 生产环境数据库
host = 'rm-2zehj3r1n60ttz9x5.mysql.rds.aliyuncs.com'
port = 3306
dbusername = 'jingbo'
password = 'shihua@123'
dbname = 'jingbo'
table_name = 'v_tbl_crude_oil_warning'
# 变量定义--测试环境
# server_host = '192.168.100.53:8080' # 内网
# # server_host = '183.242.74.28' # 外网
# login_pushreport_url = f"http://{server_host}/jingbo-dev/api/server/login"
# upload_url = f"http://{server_host}/jingbo-dev/api/analysis/reportInfo/researchUploadReportSave"
# upload_warning_url = f"http://{server_host}/jingbo-dev/api/basicBuiness/crudeOilWarning/save"
# query_data_list_item_nos_url = f"http://{server_host}/jingbo-dev/api/warehouse/dwDataItem/queryDataListItemNos"
# server_host = '10.200.32.39'
# login_pushreport_url = "http://10.200.32.39/jingbo-api/api/server/login"
# upload_url = "http://10.200.32.39/jingbo-api/api/analysis/reportInfo/researchUploadReportSave"
# upload_warning_url = "http://10.200.32.39/jingbo-api/api/basicBuiness/crudeOilWarning/save"
# query_data_list_item_nos_url = f"http://{server_host}/jingbo-api/api/warehouse/dwDataItem/queryDataListItemNos"
# # 上传数据项值
# push_data_value_list_url = f"http://{server_host}/jingbo-dev/api/dw/dataValue/pushDataValueList"
# # 上传图片报告
# push_png_report_url = f"http://{server_host}/jingbo-dev/api/analysis/reportInfo/priceForecastImg"
# push_data_value_list_url = f"http://{server_host}/jingbo-api/api/dw/dataValue/pushDataValueList"
# push_png_report_url = f"http://{server_host}/jingbo-api/api/analysis/reportInfo/priceForecastImg"
# login_data = {
# "data": {
# "account": "api_test",
# # "password": "MmVmNzNlOWI0MmY0ZDdjZGUwNzE3ZjFiMDJiZDZjZWU=", # Shihua@123456
# "password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=", # 123456
# "account": "api_dev",
# "password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=",
# "tenantHashCode": "8a4577dbd919675758d57999a1e891fe",
# "terminal": "API"
# },
@ -367,24 +238,26 @@ table_name = 'v_tbl_crude_oil_warning'
# "funcOperation": "获取token"
# }
# upload_data = {
# "funcModule": '研究报告信息',
# "funcOperation": '上传聚烯烃PP价格预测报告',
# "funcOperation": '上传原油价格预测报告',
# "data": {
# "ownerAccount": 'arui', # 报告所属用户账号
# "groupNo": '', # 用户组id
# "ownerAccount": '27663', # 报告所属用户账号 27663 - 刘小朋
# "reportType": 'OIL_PRICE_FORECAST', # 报告类型，固定为OIL_PRICE_FORECAST
# "fileName": '2000-40-5-50--100-原油指标数据.xlsx-Brent活跃合约--2024-09-06-15-01-29-预测报告.pdf', # 文件名称
# "fileName": '', # 文件名称
# "fileBase64": '', # 文件内容base64
# "categoryNo": 'yyjgycbg', # 研究报告分类编码
# "smartBusinessClassCode": 'JXTJGYCBG', # 分析报告分类编码
# "reportEmployeeCode": "E40116", # 报告人
# "reportDeptCode": "D0044", # 报告部门
# "smartBusinessClassCode": 'YCJGYCBG', # 分析报告分类编码
# "reportEmployeeCode": "E40482", # 报告人 E40482 - 管理员 0000027663 - 刘小朋
# "reportDeptCode": "002000621000", # 报告部门 - 002000621000 SH期货研究部
# "productGroupCode": "RAW_MATERIAL" # 商品分类
# }
# }
# warning_data = {
# "groupNo": '', # 用户组id
# "funcModule": '原油特征停更预警',
# "funcOperation": '原油特征停更预警',
# "data": {
@ -398,13 +271,13 @@ table_name = 'v_tbl_crude_oil_warning'
# "funcModule": "数据项",
# "funcOperation": "查询",
# "data": {
# "dateStart": "20200101",
# "dateEnd": "20241231",
# # 数据项编码,代表 PP期货 价格
# "dataItemNoList": ["MAIN_CONFT_SETTLE_PRICE"]
# "dateStart": "20150101",
# "dateEnd": "20301231",
# "dataItemNoList": ["MAIN_CONFT_SETTLE_PRICE"] # 数据项编码,代表 brent最低价和最高价
# }
# }
# push_data_value_list_data = {
# "funcModule": "数据表信息列表",
# "funcOperation": "新增",
@ -427,35 +300,183 @@ table_name = 'v_tbl_crude_oil_warning'
# ]
# }
# push_png_report_data = {
# "funcModule": '研究报告信息',
# "funcOperation": '上传聚烯烃PP价格预测报告',
# "funcModule": '聚烯烃图片报告',
# "funcOperation": '上传聚烯烃PP价格预测图片报告',
# "data": {
# "groupNo": "000161",
# "groupNo": "000211",
# "updateTime": "2024-09-06 15:01:29",
# "fileBase64": '', # 文件内容base64
# "title": '2025年8月5日日度周度预测结果',
# "billNo": '',
# "pushContent": ""
# }
# }
# 套期保值正式环境start
tqbz_login_url = "http://10.200.32.39/jbsh/api/server/login"
tqbz_login_data = {
"funcModule": "login",
"funcOperation": "login",
"data": {
"account": "api_szh",
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=",
"tenantHashCode":"7145255749fb7d61263081fca3cb1469",
"terminal": "API"
}
}
query_is_trading_day_url = "http://10.200.32.39/jbsh/api/dd/futuresCalendar/checkTradeDay"
is_trading_day_data = {
"funcModule": "检查是否交易日",
"funcOperation": "检查是否交易日",
"data": "20251010"
}
# 套期保值正式环境end
# # 八大维度数据项编码
# bdwd_items = {
# 'ciri': 'jxtppbdwdcr',
# 'benzhou': 'jxtppbdwdbz',
# 'cizhou': 'jxtppbdwdcz',
# 'gezhou': 'jxtppbdwdgz',
# 'ciyue': 'jxtppbdwdcy',
# 'cieryue': 'jxtppbdwdcey',
# 'cisanyue': 'jxtppbdwdcsany',
# 'cisiyue': 'jxtppbdwdcsiy',
# 'ciri': '251889263|FORECAST|PRICE|T01',
# 'cierri': '251889263|FORECAST|PRICE|T02',
# 'cisanri': '251889263|FORECAST|PRICE|T03',
# 'cisiri': '251889263|FORECAST|PRICE|T04',
# 'benzhou': '251889263|FORECAST|PRICE|T05',
# 'cizhou': '251889263|FORECAST|PRICE|W_01',
# 'gezhou': '251889263|FORECAST|PRICE|W_02',
# 'ciyue': '251889263|FORECAST|PRICE|M_01',
# 'cieryue': '251889263|FORECAST|PRICE|M_02',
# 'cisanyue': '251889263|FORECAST|PRICE|M_03',
# 'cisiyue': '251889263|FORECAST|PRICE|M_04',
# }
# # 报告中八大维度数据项重命名
# columnsrename = {'jxtppbdwdbz': '本周', 'jxtppbdwdcey': '次二月', 'jxtppbdwdcr': '次日', 'jxtppbdwdcsiy': '次四月',
# 'jxtppbdwdcsany': '次三月', 'jxtppbdwdcy': '次月', 'jxtppbdwdcz': '次周', 'jxtppbdwdgz': '隔周', }
# columnsrename = {'251889263|FORECAST|PRICE|T05': '本周', '251889263|FORECAST|PRICE|M_02': '次二月', '251889263|FORECAST|PRICE|T01': '次日', '251889263|FORECAST|PRICE|M_04': '次四月',
# '251889263|FORECAST|PRICE|M_03': '次三月', '251889263|FORECAST|PRICE|M_01': '次月', '251889263|FORECAST|PRICE|W_01': '次周', '251889263|FORECAST|PRICE|W_02': '隔周', }
# # 生产环境数据库
# host = 'rm-2zehj3r1n60ttz9x5.mysql.rds.aliyuncs.com'
# port = 3306
# dbusername = 'jingbo'
# password = 'shihua@123'
# dbname = 'jingbo'
# table_name = 'v_tbl_crude_oil_warning'
# 变量定义--测试环境
server_host = '192.168.100.53:8080' # 内网
# server_host = '183.242.74.28' # 外网
login_pushreport_url = f"http://{server_host}/jingbo-dev/api/server/login"
upload_url = f"http://{server_host}/jingbo-dev/api/analysis/reportInfo/researchUploadReportSave"
upload_warning_url = f"http://{server_host}/jingbo-dev/api/basicBuiness/crudeOilWarning/save"
query_data_list_item_nos_url = f"http://{server_host}/jingbo-dev/api/warehouse/dwDataItem/queryDataListItemNos"
# 上传数据项值
push_data_value_list_url = f"http://{server_host}/jingbo-dev/api/dw/dataValue/pushDataValueList"
# 上传图片报告
push_png_report_url = f"http://{server_host}/jingbo-dev/api/analysis/reportInfo/priceForecastImg"
login_data = {
"data": {
"account": "api_test",
# "password": "MmVmNzNlOWI0MmY0ZDdjZGUwNzE3ZjFiMDJiZDZjZWU=", # Shihua@123456
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=", # 123456
"tenantHashCode": "8a4577dbd919675758d57999a1e891fe",
"terminal": "API"
},
"funcModule": "API",
"funcOperation": "获取token"
}
upload_data = {
"funcModule": '研究报告信息',
"funcOperation": '上传聚烯烃PP价格预测报告',
"data": {
"ownerAccount": 'arui', # 报告所属用户账号
"reportType": 'OIL_PRICE_FORECAST', # 报告类型，固定为OIL_PRICE_FORECAST
"fileName": '2000-40-5-50--100-原油指标数据.xlsx-Brent活跃合约--2024-09-06-15-01-29-预测报告.pdf', # 文件名称
"fileBase64": '', # 文件内容base64
"categoryNo": 'yyjgycbg', # 研究报告分类编码
"smartBusinessClassCode": 'JXTJGYCBG', # 分析报告分类编码
"reportEmployeeCode": "E40116", # 报告人
"reportDeptCode": "D0044", # 报告部门
"productGroupCode": "RAW_MATERIAL" # 商品分类
}
}
warning_data = {
"funcModule": '原油特征停更预警',
"funcOperation": '原油特征停更预警',
"data": {
'WARNING_TYPE_NAME': '特征数据停更预警',
'WARNING_CONTENT': '',
'WARNING_DATE': ''
}
}
query_data_list_item_nos_data = {
"funcModule": "数据项",
"funcOperation": "查询",
"data": {
"dateStart": "20200101",
"dateEnd": "20241231",
# 数据项编码,代表 PP期货 价格
"dataItemNoList": ["MAIN_CONFT_SETTLE_PRICE"]
}
}
push_data_value_list_data = {
"funcModule": "数据表信息列表",
"funcOperation": "新增",
"data": [
{"dataItemNo": "91230600716676129",
"dataDate": "20230113",
"dataStatus": "add",
"dataValue": 100.11
},
{"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
"dataDate": "20230113",
"dataStatus": "add",
"dataValue": 100.55
},
{"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
"dataDate": "20230113",
"dataStatus": "add",
"dataValue": 100.55
}
]
}
push_png_report_data = {
"funcModule": '研究报告信息',
"funcOperation": '上传聚烯烃PP价格预测报告',
"data": {
"groupNo": "000161",
"updateTime": "2024-09-06 15:01:29",
"fileBase64": '', # 文件内容base64
"title": '2025年8月5日日度周度预测结果',
"billNo": '',
"pushContent": "2025年8月5日日度周度预测结果"
}
}
# 八大维度数据项编码
bdwd_items = {
'ciri': 'jxtppbdwdcr',
'benzhou': 'jxtppbdwdbz',
'cizhou': 'jxtppbdwdcz',
'gezhou': 'jxtppbdwdgz',
'ciyue': 'jxtppbdwdcy',
'cieryue': 'jxtppbdwdcey',
'cisanyue': 'jxtppbdwdcsany',
'cisiyue': 'jxtppbdwdcsiy',
}
# 报告中八大维度数据项重命名
columnsrename = {'jxtppbdwdbz': '本周', 'jxtppbdwdcey': '次二月', 'jxtppbdwdcr': '次日', 'jxtppbdwdcsiy': '次四月',
'jxtppbdwdcsany': '次三月', 'jxtppbdwdcy': '次月', 'jxtppbdwdcz': '次周', 'jxtppbdwdgz': '隔周', }
# 北京环境数据库
host = '192.168.101.27'

379
config_juxiting_zhoudu.py
View File

@ -169,23 +169,185 @@ ClassifyId = 1161
# 变量定义--线上环境
server_host = '10.200.32.39'
login_pushreport_url = "http://10.200.32.39/jingbo-api/api/server/login"
upload_url = "http://10.200.32.39/jingbo-api/api/analysis/reportInfo/researchUploadReportSave"
upload_warning_url = "http://10.200.32.39/jingbo-api/api/basicBuiness/crudeOilWarning/save"
query_data_list_item_nos_url = f"http://{server_host}/jingbo-api/api/warehouse/dwDataItem/queryDataListItemNos"
# 上传数据项值
push_data_value_list_url = f"http://{server_host}/jingbo-api/api/dw/dataValue/pushDataValueList"
# 上传停更数据到市场信息平台
push_waring_data_value_list_url = f"http://{server_host}/jingbo-api/api/basicBuiness/crudeOilWarning/crudeSaveOrupdate"
# 获取预警数据中取消订阅指标ID
get_waring_data_value_list_url = f"http://{server_host}/jingbo-api/api/basicBuiness/crudeOilWarning/dataList"
# server_host = '10.200.32.39'
# login_pushreport_url = "http://10.200.32.39/jingbo-api/api/server/login"
# upload_url = "http://10.200.32.39/jingbo-api/api/analysis/reportInfo/researchUploadReportSave"
# upload_warning_url = "http://10.200.32.39/jingbo-api/api/basicBuiness/crudeOilWarning/save"
# query_data_list_item_nos_url = f"http://{server_host}/jingbo-api/api/warehouse/dwDataItem/queryDataListItemNos"
# # 上传数据项值
# push_data_value_list_url = f"http://{server_host}/jingbo-api/api/dw/dataValue/pushDataValueList"
# # 上传停更数据到市场信息平台
# push_waring_data_value_list_url = f"http://{server_host}/jingbo-api/api/basicBuiness/crudeOilWarning/crudeSaveOrupdate"
# # 获取预警数据中取消订阅指标ID
# get_waring_data_value_list_url = f"http://{server_host}/jingbo-api/api/basicBuiness/crudeOilWarning/dataList"
# login_data = {
# "data": {
# "account": "api_dev",
# "password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=",
# "tenantHashCode": "8a4577dbd919675758d57999a1e891fe",
# "terminal": "API"
# },
# "funcModule": "API",
# "funcOperation": "获取token"
# }
# upload_data = {
# "funcModule": '研究报告信息',
# "funcOperation": '上传聚烯烃PP价格预测报告',
# "data": {
# "groupNo": '000211', # 用户组编号
# "ownerAccount": '36541', # 报告所属用户账号  36541 - 贾青雪
# "reportType": 'OIL_PRICE_FORECAST', # 报告类型，固定为OIL_PRICE_FORECAST
# "fileName": '', # 文件名称
# "fileBase64": '', # 文件内容base64
# "categoryNo": 'jxtjgycbg', # 研究报告分类编码
# "smartBusinessClassCode": 'JXTJGYCBG', # 分析报告分类编码
# "reportEmployeeCode": "E40482", # 报告人  E40482  - 管理员  0000027663 - 刘小朋  
# "reportDeptCode": "JXTJGYCBG", # 报告部门 - 002000621000  SH期货研究部  
# "productGroupCode": "RAW_MATERIAL" # 商品分类
# }
# }
# warning_data = {
# "funcModule": '原油特征停更预警',
# "funcOperation": '原油特征停更预警',
# "data": {
# "groupNo": "000211",
# 'WARNING_TYPE_NAME': '特征数据停更预警',
# 'WARNING_CONTENT': '',
# 'WARNING_DATE': ''
# }
# }
# query_data_list_item_nos_data = {
# "funcModule": "数据项",
# "funcOperation": "查询",
# "data": {
# "dateStart": "20200101",
# "dateEnd": "",
# # 数据项编码,代表 PP期货 价格
# "dataItemNoList": ["MAIN_CONFT_SETTLE_PRICE"]
# }
# }
# push_data_value_list_data = {
# "funcModule": "数据表信息列表",
# "funcOperation": "新增",
# "data": [
# {"dataItemNo": "91230600716676129",
# "dataDate": "20230113",
# "dataStatus": "add",
# "dataValue": 100.11
# },
# {"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
# "dataDate": "20230113",
# "dataStatus": "add",
# "dataValue": 100.55
# },
# {"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
# "dataDate": "20230113",
# "dataStatus": "add",
# "dataValue": 100.55
# }
# ]
# }
# push_waring_data_value_list_data = {
# "data": {
# "crudeOilWarningDtoList": [
# {
# "lastUpdateDate": "20240501",
# "updateSuspensionCycle": 1,
# "dataSource": "9",
# "frequency": "1",
# "indicatorName": "美元指数",
# "indicatorId": "myzs001",
# "warningDate": "2024-05-13"
# }
# ],
# "dataSource": "9"
# },
# "funcModule": "商品数据同步",
# "funcOperation": "同步"
# }
# get_waring_data_value_list_data = {
# "data": "9", "funcModule": "商品数据同步", "funcOperation": "同步"}
# 套期保值正式环境start
tqbz_login_url = "http://10.200.32.39/jbsh/api/server/login"
tqbz_login_data = {
"funcModule": "login",
"funcOperation": "login",
"data": {
"account": "api_szh",
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=",
"tenantHashCode":"7145255749fb7d61263081fca3cb1469",
"terminal": "API"
}
}
query_is_trading_day_url = "http://10.200.32.39/jbsh/api/dd/futuresCalendar/checkTradeDay"
is_trading_day_data = {
"funcModule": "检查是否交易日",
"funcOperation": "检查是否交易日",
"data": "20251010"
}
# 套期保值正式环境end
# # 八大维度数据项编码
# bdwd_items = {
# 'ciri': '251889263|FORECAST|PRICE|T01',
# 'cierri': '251889263|FORECAST|PRICE|T02',
# 'cisanri': '251889263|FORECAST|PRICE|T03',
# 'cisiri': '251889263|FORECAST|PRICE|T04',
# 'benzhou': '251889263|FORECAST|PRICE|T05',
# 'cizhou': '251889263|FORECAST|PRICE|W_01',
# 'gezhou': '251889263|FORECAST|PRICE|W_02',
# 'ciyue': '251889263|FORECAST|PRICE|M_01',
# 'cieryue': '251889263|FORECAST|PRICE|M_02',
# 'cisanyue': '251889263|FORECAST|PRICE|M_03',
# 'cisiyue': '251889263|FORECAST|PRICE|M_04',
# }
# # 生产环境数据库
# host = 'rm-2zehj3r1n60ttz9x5.mysql.rds.aliyuncs.com'
# port = 3306
# dbusername = 'jingbo'
# password = 'shihua@123'
# dbname = 'jingbo'
# table_name = 'v_tbl_crude_oil_warning'
# 变量定义--测试环境
server_host = '192.168.100.53' # 内网
# server_host = '183.242.74.28' # 外网
login_pushreport_url = f"http://{server_host}:8080/jingbo-dev/api/server/login"
# 上传报告
upload_url = f"http://{server_host}:8080/jingbo-dev/api/analysis/reportInfo/researchUploadReportSave"
# 停更预警
upload_warning_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/save"
# 查询数据项编码
query_data_list_item_nos_url = f"http://{server_host}:8080/jingbo-dev/api/warehouse/dwDataItem/queryDataListItemNos"
# 上传数据项值
push_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/dw/dataValue/pushDataValueList"
# 上传停更数据到市场信息平台
push_waring_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/crudeSaveOrupdate"
# 获取预警数据中取消订阅指标ID
get_waring_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/dataList"
login_data = {
"data": {
"account": "api_dev",
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=",
"account": "api_test",
# "password": "MmVmNzNlOWI0MmY0ZDdjZGUwNzE3ZjFiMDJiZDZjZWU=", # Shihua@123456
"password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=", # 123456
"tenantHashCode": "8a4577dbd919675758d57999a1e891fe",
"terminal": "API"
},
@ -193,29 +355,29 @@ login_data = {
"funcOperation": "获取token"
}
upload_data = {
"funcModule": '研究报告信息',
"funcOperation": '上传聚烯烃PP价格预测报告',
"data": {
"groupNo": '000211', # 用户组编号
"ownerAccount": '36541', # 报告所属用户账号  36541 - 贾青雪
"groupNo": "000127",
"ownerAccount": 'arui', # 报告所属用户账号
"reportType": 'OIL_PRICE_FORECAST', # 报告类型，固定为OIL_PRICE_FORECAST
"fileName": '', # 文件名称
"fileName": '2000-40-5-50--100-原油指标数据.xlsx-Brent活跃合约--2024-09-06-15-01-29-预测报告.pdf', # 文件名称
"fileBase64": '', # 文件内容base64
"categoryNo": 'jxtjgycbg', # 研究报告分类编码
"categoryNo": 'yyjgycbg', # 研究报告分类编码
"smartBusinessClassCode": 'JXTJGYCBG', # 分析报告分类编码
"reportEmployeeCode": "E40482", # 报告人  E40482  - 管理员  0000027663 - 刘小朋  
"reportDeptCode": "JXTJGYCBG", # 报告部门 - 002000621000  SH期货研究部  
"productGroupCode": "RAW_MATERIAL" # 商品分类
"reportEmployeeCode": "E40116", # 报告人
"reportDeptCode": "D0044", # 报告部门
"productGroupCode": "RAW_MATERIAL" # 商品分类
}
}
# 已弃用
warning_data = {
"funcModule": '原油特征停更预警',
"funcOperation": '原油特征停更预警',
"data": {
"groupNo": "000211",
"groupNo": "000127",
'WARNING_TYPE_NAME': '特征数据停更预警',
'WARNING_CONTENT': '',
'WARNING_DATE': ''
@ -233,7 +395,6 @@ query_data_list_item_nos_data = {
}
}
push_data_value_list_data = {
"funcModule": "数据表信息列表",
"funcOperation": "新增",
@ -281,170 +442,26 @@ get_waring_data_value_list_data = {
"data": "9", "funcModule": "商品数据同步", "funcOperation": "同步"}
# 八大维度数据项编码
bdwd_items = {
'ciri': '251889263|FORECAST|PRICE|T01',
'cierri': '251889263|FORECAST|PRICE|T02',
'cisanri': '251889263|FORECAST|PRICE|T03',
'cisiri': '251889263|FORECAST|PRICE|T04',
'benzhou': '251889263|FORECAST|PRICE|T05',
'cizhou': '251889263|FORECAST|PRICE|W_01',
'gezhou': '251889263|FORECAST|PRICE|W_02',
'ciyue': '251889263|FORECAST|PRICE|M_01',
'cieryue': '251889263|FORECAST|PRICE|M_02',
'cisanyue': '251889263|FORECAST|PRICE|M_03',
'cisiyue': '251889263|FORECAST|PRICE|M_04',
'ciri': 'jxtppbdwdcr',
'benzhou': 'jxtppbdwdbz',
'cizhou': 'jxtppbdwdcz',
'gezhou': 'jxtppbdwdgz',
'ciyue': 'jxtppbdwdcy',
'cieryue': 'jxtppbdwdcey',
'cisanyue': 'jxtppbdwdcsany',
'cisiyue': 'jxtppbdwdcsiy',
}
# # 生产环境数据库
host = 'rm-2zehj3r1n60ttz9x5.mysql.rds.aliyuncs.com'
port = 3306
dbusername = 'jingbo'
password = 'shihua@123'
dbname = 'jingbo'
table_name = 'v_tbl_crude_oil_warning'
# 变量定义--测试环境
# server_host = '192.168.100.53' # 内网
# # server_host = '183.242.74.28' # 外网
# login_pushreport_url = f"http://{server_host}:8080/jingbo-dev/api/server/login"
# # 上传报告
# upload_url = f"http://{server_host}:8080/jingbo-dev/api/analysis/reportInfo/researchUploadReportSave"
# # 停更预警
# upload_warning_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/save"
# # 查询数据项编码
# query_data_list_item_nos_url = f"http://{server_host}:8080/jingbo-dev/api/warehouse/dwDataItem/queryDataListItemNos"
# # 上传数据项值
# push_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/dw/dataValue/pushDataValueList"
# # 上传停更数据到市场信息平台
# push_waring_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/crudeSaveOrupdate"
# # 获取预警数据中取消订阅指标ID
# get_waring_data_value_list_url = f"http://{server_host}:8080/jingbo-dev/api/basicBuiness/crudeOilWarning/dataList"
# login_data = {
# "data": {
# "account": "api_test",
# # "password": "MmVmNzNlOWI0MmY0ZDdjZGUwNzE3ZjFiMDJiZDZjZWU=", # Shihua@123456
# "password": "ZTEwYWRjMzk0OWJhNTlhYmJlNTZlMDU3ZjIwZjg4M2U=", # 123456
# "tenantHashCode": "8a4577dbd919675758d57999a1e891fe",
# "terminal": "API"
# },
# "funcModule": "API",
# "funcOperation": "获取token"
# }
# upload_data = {
# "funcModule": '研究报告信息',
# "funcOperation": '上传聚烯烃PP价格预测报告',
# "data": {
# "groupNo": "000127",
# "ownerAccount": 'arui', # 报告所属用户账号
# "reportType": 'OIL_PRICE_FORECAST', # 报告类型，固定为OIL_PRICE_FORECAST
# "fileName": '2000-40-5-50--100-原油指标数据.xlsx-Brent活跃合约--2024-09-06-15-01-29-预测报告.pdf', # 文件名称
# "fileBase64": '', # 文件内容base64
# "categoryNo": 'yyjgycbg', # 研究报告分类编码
# "smartBusinessClassCode": 'JXTJGYCBG', # 分析报告分类编码
# "reportEmployeeCode": "E40116", # 报告人
# "reportDeptCode": "D0044", # 报告部门
# "productGroupCode": "RAW_MATERIAL" # 商品分类
# }
# }
# # 已弃用
# warning_data = {
# "funcModule": '原油特征停更预警',
# "funcOperation": '原油特征停更预警',
# "data": {
# "groupNo": "000127",
# 'WARNING_TYPE_NAME': '特征数据停更预警',
# 'WARNING_CONTENT': '',
# 'WARNING_DATE': ''
# }
# }
# query_data_list_item_nos_data = {
# "funcModule": "数据项",
# "funcOperation": "查询",
# "data": {
# "dateStart": "20200101",
# "dateEnd": "",
# # 数据项编码,代表 PP期货 价格
# "dataItemNoList": ["MAIN_CONFT_SETTLE_PRICE"]
# }
# }
# push_data_value_list_data = {
# "funcModule": "数据表信息列表",
# "funcOperation": "新增",
# "data": [
# {"dataItemNo": "91230600716676129",
# "dataDate": "20230113",
# "dataStatus": "add",
# "dataValue": 100.11
# },
# {"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
# "dataDate": "20230113",
# "dataStatus": "add",
# "dataValue": 100.55
# },
# {"dataItemNo": "91230600716676129P|ETHYL_BEN|CAPACITY",
# "dataDate": "20230113",
# "dataStatus": "add",
# "dataValue": 100.55
# }
# ]
# }
# push_waring_data_value_list_data = {
# "data": {
# "crudeOilWarningDtoList": [
# {
# "lastUpdateDate": "20240501",
# "updateSuspensionCycle": 1,
# "dataSource": "9",
# "frequency": "1",
# "indicatorName": "美元指数",
# "indicatorId": "myzs001",
# "warningDate": "2024-05-13"
# }
# ],
# "dataSource": "9"
# },
# "funcModule": "商品数据同步",
# "funcOperation": "同步"
# }
# get_waring_data_value_list_data = {
# "data": "9", "funcModule": "商品数据同步", "funcOperation": "同步"}
# # 八大维度数据项编码
# bdwd_items = {
# 'ciri': 'jxtppbdwdcr',
# 'benzhou': 'jxtppbdwdbz',
# 'cizhou': 'jxtppbdwdcz',
# 'gezhou': 'jxtppbdwdgz',
# 'ciyue': 'jxtppbdwdcy',
# 'cieryue': 'jxtppbdwdcey',
# 'cisanyue': 'jxtppbdwdcsany',
# 'cisiyue': 'jxtppbdwdcsiy',
# }
# 北京环境数据库
# host = '192.168.101.27'
# port = 3306
# dbusername = 'root'
# password = '123456'
# dbname = 'jingbo_test'
# table_name = 'v_tbl_crude_oil_warning'
host = '192.168.101.27'
port = 3306
dbusername = 'root'
password = '123456'
dbname = 'jingbo_test'
table_name = 'v_tbl_crude_oil_warning'
DEFAULT_CONFIG = {
'feature_factor_frequency': 'D',

29
juxiting_push_png_report.py
View File

@ -106,6 +106,12 @@ def push_png_report():
data['data']['fileBase64'] = base64_data
data['data']['billNo'] = str(time.time())
if png_report_file == 'pp_zhouducorrelation.png':
data['data']['pushContent'] = f'{end_time}PP期货日、周维度预测价格走势'
else:
data['data']['pushContent'] = f'{end_time}PP期货月维度预测价格走势'
# data['data']['pushContent'] = f'{end_time}PP期货价格预测'
pngreportdata = push_png_report_to_market(data)
logger.info(f'{png_report_file}推送图片报告到钉钉成功{pngreportdata}')
except Exception as e:
@ -116,10 +122,27 @@ if __name__ == '__main__':
# 图片报告
try:
logger.info('图片报告ing')
global_config['end_time'] = '2025-08-14'
global_config['end_time'] = '2025-08-01'
# previous_trading_day = (pd.Timestamp(global_config['end_time']) -
# pd.tseries.offsets.BusinessDay(1)).strftime('%Y-%m-%d')
# global_config['end_time'] = previous_trading_day
# 交易日检查
# 将 end_time 转换为 YYYYMMDD 格式
end_time_str = global_config['end_time'].replace('-', '')
logger.info(f"开始检查预测日期 {global_config['end_time']} 是否为交易日")
# 使用 global_config 中的交易日检查参数
is_trading = check_trading_day(end_time_str, global_config)
if not is_trading:
logger.info(f"预测日期 {global_config['end_time']} 不是交易日，跳过图片报告任务")
exit(0)
else:
logger.info(f"预测日期 {global_config['end_time']} 是交易日，开始执行图片报告任务")
pp_bdwd_png(global_config=global_config)
logger.info('图片报告end')
except Exception as e:
logger.info(f'图片报告失败：{e}')
# time.sleep(5)
# push_png_report()
time.sleep(5)
push_png_report()

142
lib/dataread.py
View File

@ -80,6 +80,8 @@ global_config = {
'upload_headers': None, # 上传请求头
'upload_warning_url': None, # 预警数据上传地址
'upload_warning_data': None, # 预警数据结构
'upload_report_audit_url': None, # 报告审核地址
'upload_report_audit_data': None, # 报告审核数据结构
# 报告上传
'upload_data': None, # 报告数据结构
@ -95,6 +97,12 @@ global_config = {
'push_waring_data_value_list_url': None,
'push_waring_data_value_list_data': None,
# 套期保值api
"tqbz_login_url":None,
"tqbz_login_data":None,
'query_is_trading_day_url': None,
'is_trading_day_data': None,
# 字段映射
'offsite_col': None, # 站点字段
'avg_col': None, # 平均值字段
@ -231,7 +239,7 @@ def get_head_auth_report():
f'url:{config.login_pushreport_url},login_data:{config.login_data}')
# 发送 POST 请求到登录 URL，携带登录数据
login_res = requests.post(url=config.login_pushreport_url,
json=config.login_data, timeout=(3, 30))
json=config.login_data, timeout=(10, 30))
# 将响应内容转换为 JSON 格式
text = json.loads(login_res.text)
@ -272,7 +280,7 @@ def upload_report_data(token, upload_data):
# 发送POST请求，上传报告数据
upload_res = requests.post(
url=config.upload_url, headers=headers, json=upload_data, timeout=(3, 15))
url=config.upload_url, headers=headers, json=upload_data, timeout=(10, 15))
# 将响应内容转换为 JSON 格式
upload_res = json.loads(upload_res.text)
@ -319,7 +327,7 @@ def upload_warning_data(warning_data):
# 发送POST请求，上传预警数据
upload_res = requests.post(
url=config.upload_warning_url, headers=headers, json=config.warning_data, timeout=(3, 15))
url=config.upload_warning_url, headers=headers, json=config.warning_data, timeout=(10, 15))
# 如果上传成功，返回响应对象
if upload_res:
@ -900,21 +908,21 @@ def datachuli(df_zhibiaoshuju, df_zhibiaoliebiao, datecol='date', end_time='', y
# 删除全为空值的列
df = df.dropna(axis=1, how='all')
# 删除开始时间没有数据的列
config.logger.info(f'删除开始时间没有数据的列前数据量：{df.shape}')
for col in df.columns:
if col == 'ds': # 跳过 'ds' 列
continue
# 找到第一个非空值的索引
first_valid_index = df[col].first_valid_index()
if first_valid_index is not None:
# 判断对应的 'ds' 是否大于 start_date
if df.loc[first_valid_index, 'ds'] > start_date:
df.drop(columns=[col], inplace=True)
config.logger.info(
f'删除开始时间没有数据的列：{col},第一条数据日期为：{df.loc[first_valid_index, "ds"]}')
# # 删除开始时间没有数据的列
# config.logger.info(f'删除开始时间没有数据的列前数据量：{df.shape}')
# for col in df.columns:
# if col in ['ds','Brentzgj','Brentzdj']: # 跳过 'ds' 列
# continue
# # 找到第一个非空值的索引
# first_valid_index = df[col].first_valid_index()
# if first_valid_index is not None:
# # 判断对应的 'ds' 是否大于 start_date
# if df.loc[first_valid_index, 'ds'] > start_date:
# df.drop(columns=[col], inplace=True)
# config.logger.info(
# f'删除开始时间没有数据的列：{col},第一条数据日期为：{df.loc[first_valid_index, "ds"]}')
config.logger.info(f'删除开始时间没有数据的列后数据量：{df.shape}')
# config.logger.info(f'删除开始时间没有数据的列后数据量：{df.shape}')
# 获取start_year年到end_time的数据
df = df[df['ds'].dt.year >= config.start_year]
@ -947,6 +955,8 @@ def datachuli(df_zhibiaoshuju, df_zhibiaoliebiao, datecol='date', end_time='', y
if is_timefurture:
df = addtimecharacteristics(df=df, dataset=dataset)
if config.freq == 'WW':
# 自定义周数据
# 按weekofmothe分组取均值得到新的数据
@ -994,6 +1004,14 @@ def datachuli(df_zhibiaoshuju, df_zhibiaoliebiao, datecol='date', end_time='', y
# kdj指标
if add_kdj:
df = calculate_kdj(df)
# 衍生预测目标近30日的中位数，最大，最小
if config.freq == 'M':
df['y_30d_median'] = df['y'].rolling(window=30).median()
df['y_30d_max'] = df['y'].rolling(window=30).max()
df['y_30d_min'] = df['y'].rolling(window=30).min()
# 保存填充后的数据
df.to_csv(os.path.join(
dataset, '处理后的特征数据.csv'), index=False)
@ -1042,6 +1060,18 @@ def zhoududatachuli(df_zhibiaoshuju, df_zhibiaoliebiao, datecol='date', end_time
config.logger.info(f'删除两月不更新特征后数据量：{df.shape}')
# 在重采样之前：优先将最高/最低价从原始数据拼接进来，并进行缺失值填充
try:
for col in ['Brentzgj', 'Brentzdj']:
if col in df_zhibiaoshuju.columns and col not in df.columns:
df = pd.merge(df, df_zhibiaoshuju[['ds', col]], on='ds', how='left')
# 对已存在的高低价列进行前后向填充，确保重采样时不产生 NaN
for col in ['Brentzgj', 'Brentzdj']:
if col in df.columns:
df[col] = df[col].ffill().bfill()
except Exception as e:
config.logger.info(f'重采样前拼接高低价失败：{e}')
if config.freq == 'W':
# 按周取样
df = df.resample('W', on='ds').mean().reset_index()
@ -1435,6 +1465,12 @@ class Config:
def upload_warning_data(self): return global_config['upload_warning_data']
@property
def warning_data(self): return global_config['warning_data']
@property
def upload_report_audit_url(self): return global_config['upload_report_audit_url']
@property
def upload_report_audit_data(self): return global_config['upload_report_audit_data']
# 查询接口
@property
@ -2454,7 +2490,7 @@ def get_market_data(end_time, df):
headers = {"Authorization": token}
config.logger.info('获取数据中...')
items_res = requests.post(url=config.query_data_list_item_nos_url, headers=headers,
json=config.query_data_list_item_nos_data, timeout=(3, 35))
json=config.query_data_list_item_nos_data, timeout=(10, 35))
json_data = json.loads(items_res.text)
df3 = pd.DataFrame(json_data['data'])
# 按照dataItemNo 分组 得到多个dataframe ，最后根据dataDate merge 成一个dataframe
@ -2491,10 +2527,21 @@ def push_png_report_to_market(data):
config.logger.info(f'推送图片报告URL：{config.push_png_report_url}')
# config.logger.info(f'推送图片报告数据：{data}')
items_res = requests.post(url=config.push_png_report_url, headers=headers,
json=data, timeout=(3, 35))
json=data, timeout=(10, 35))
json_data = json.loads(items_res.text)
config.logger.info(f"推送图片报告结果:{json_data}")
return json_data
# 发送请求
# headers = {"Authorization": token}
# config.logger.info('推送图片报告中...')
# config.logger.info(f'推送图片报告URL：{config.upload_report_audit_url}')
# # config.logger.info(f'推送图片报告数据：{data}')
# items_res = requests.post(url=config.upload_report_audit_url, headers=headers,
# json=data, timeout=(10, 35))
# json_data = json.loads(items_res.text)
# config.logger.info(f"推送图片报告结果:{json_data}")
# return json_data
def push_market_data(data):
@ -2532,7 +2579,7 @@ def push_market_data(data):
config.logger.info(f'上传数据数据：{config.push_data_value_list_data}')
items_res = requests.post(url=config.push_data_value_list_url, headers=headers,
json=config.push_data_value_list_data, timeout=(3, 35))
json=config.push_data_value_list_data, timeout=(10, 35))
json_data = json.loads(items_res.text)
config.logger.info(f"上传结果:{json_data}")
return json_data
@ -2572,7 +2619,7 @@ def push_waring_market_data(data, dataSource=8):
headers = {"Authorization": token}
config.logger.info('上传数据中...')
items_res = requests.post(url=config.push_waring_data_value_list_url, headers=headers,
json=config.push_waring_data_value_list_data, timeout=(3, 35))
json=config.push_waring_data_value_list_data, timeout=(10, 35))
json_data = json.loads(items_res.text)
config.logger.info(f"上传结果:{json_data}")
return json_data
@ -2606,7 +2653,7 @@ def get_waring_data():
headers = {"Authorization": token}
config.logger.info('获取取消订阅指标ID中...')
items_res = requests.post(url=config.get_waring_data_value_list_url, headers=headers,
json=config.get_waring_data_value_list_data, timeout=(3, 35))
json=config.get_waring_data_value_list_data, timeout=(10, 35))
json_data = json.loads(items_res.text)
json_data = json_data['data']
quxiaodingyueidlist = []
@ -2696,8 +2743,7 @@ def get_baichuan_data(baichuanidnamedict):
print(sql)
# 获取查询结果
results = db.execute_query(sql)
df = pd.DataFrame(results, columns=[
'BAICHUAN_ID', 'DATA_DATE', 'DATA_VALUE'])
df = pd.DataFrame(results, columns=['BAICHUAN_ID', 'DATA_DATE', 'DATA_VALUE'])
df['BAICHUAN_ID'] = df['BAICHUAN_ID'].astype('string')
df.to_csv(os.path.join(config.dataset, '百川数据.csv'), index=False)
@ -2746,7 +2792,7 @@ def get_bdwd_predict_data():
# 发送POST请求，上传预警数据
respose = requests.post(
url=config.upload_warning_url, headers=headers, json=query_data_list_item_nos_data, timeout=(3, 15))
url=config.upload_warning_url, headers=headers, json=query_data_list_item_nos_data, timeout=(10, 15))
# 如果上传成功，返回响应对象
if respose:
@ -2779,7 +2825,7 @@ def get_bdwd_predict_data():
headers = {"Authorization": token}
config.logger.info('获取数据中...')
items_res = requests.post(url=config.query_data_list_item_nos_url, headers=headers,
json=query_data_list_item_nos_data, timeout=(3, 35))
json=query_data_list_item_nos_data, timeout=(10, 35))
json_data = json.loads(items_res.text)
config.logger.info(f"获取到的数据:{json_data}")
df3 = pd.DataFrame(json_data['data'])
@ -2811,3 +2857,47 @@ def get_bdwd_predict_data():
print(df2)
df2 = df2[existing_columns]
return df2
def check_trading_day(date_str, global_config):
"""
检查指定日期是否为交易日
Args:
date_str: 日期字符串，格式为 'YYYYMMDD'
global_config: 全局配置对象，包含数据库配置信息
Returns:
bool: True表示是交易日，False表示不是交易日
"""
try:
# 获取数据库连接配置
db_config = global_config.get('db_mysql')
if not db_config:
global_config['logger'].error('未找到数据库配置(db_mysql)，默认视为交易日')
return True
# 确保数据库连接是活跃的
if not db_config.is_connected():
db_config.connect()
# 使用提供的SQL语句查询交易日，直接使用传入的日期格式
sql = f"""
SELECT vtdcd.CALENDAR_DAY
FROM jingbo_test.v_tbl_dd_calendar_day AS vtdcd
WHERE (CALENDAR_NO = 'pp') AND (DELETE_FLAG = '0') AND (CALENDAR_DAY = '{date_str}')
"""
# 执行查询
result = db_config.execute_query(sql)
# 如果查询结果不为空，则该日期是交易日
is_trading = len(result) > 0 if result is not None else False
global_config['logger'].info(f'日期 {date_str} 是否为交易日: {is_trading}')
return is_trading
except Exception as e:
global_config['logger'].error(f'检查交易日时发生错误: {str(e)}，默认视为交易日')
# 数据库查询失败时，默认视为交易日
return True

433
lib/tools.py
View File

@ -862,7 +862,264 @@ def convert_df_to_pydantic_pp(df_predict, model_id_name_dict, global_config):
return results
def find_best_models_yuanyou(date='', global_config=None):
# 说明：为各价格维度选择“最佳模型”。
# 思路：按维度构造真实价基准（上一交易日/参考周均价/参考月均价），
# 比较各模型该维度预测与基准的绝对误差（可选按趋势过滤），
# 选误差最小的模型，并返回模型 id、名称、预测值与对应日期。
best_models = {}
model_id_name_dict = get_model_id_name_dict(global_config=global_config)
# 处理日期输入
if not date:
date = datetime.datetime.now().strftime('%Y-%m-%d')
else:
try:
date = datetime.datetime.strptime(
date, '%Y-%m-%d').strftime('%Y-%m-%d')
except ValueError:
global_config['logger'].error(
f"日期格式错误，期望格式为 '%Y-%m-%d'，实际输入: {date}")
return best_models
current_date = datetime.datetime.strptime(date, '%Y-%m-%d')
# 上一交易日日期 Last trading day
last_trading_day = pd.Timestamp(date) - pd.tseries.offsets.BusinessDay(1)
last_trading_day_str = last_trading_day.strftime('%Y-%m-%d')
# 计算date对应月的一日
first_day_of_month = current_date.replace(day=1)
# 计算date对应周的周一
date_monday = current_date - \
datetime.timedelta(days=current_date.weekday())
# 获取真实价格数据
try:
true_price = pd.read_csv('yuanyoudataset/指标数据.csv')[['ds', 'y']]
except FileNotFoundError:
global_config['logger'].error(
f"未找到文件: {os.path.join(global_config['dataset'], '指标数据.csv')}")
return best_models
# 计算六月前的年月
year, month = map(int, date.split('-')[:2])
if month <= 6:
year -= 1
month = 12
else:
month -= 6
tb = 'v_tbl_predict_prediction_results'
sql = f'select * from {tb} where data_date >= \'{year}-{month}-01\''
# 数据库查询对应日期的预测值
predictresult = global_config['db_mysql'].execute_query(sql)
if not predictresult:
global_config['logger'].info('没有预测结果')
return best_models
df = pd.DataFrame(predictresult)[
['data_date', 'model_id'] + global_config['price_columns']]
global_config['logger'].info(f'预测结果数量：{df.shape}')
global_config['logger'].info(
f'预测结果日期范围：{df["data_date"].min()} 到 {df["data_date"].max()}')
def query_predict_result(date, model_id, global_config, wd):
tb = 'v_tbl_predict_prediction_results'
sql = f'select {wd} from {tb} where data_date = \'{date}\' and model_id = {model_id}'
predictresult = global_config['db_mysql'].execute_query(sql)
if not predictresult:
global_config['logger'].info('没有预测结果')
return None
predictresult = float(predictresult[0][wd])
return predictresult
def calculate_best_model(price, trend, weektrueprice=None, monthtrueprice=None):
"""
计算最佳模型：
- price：候选模型在目标维度的预测（可聚合为周/月均）
- trend：当需要方向过滤时传入（与真实价差的符号一致）
- weektrueprice/monthtrueprice：周/月维度基准均值
返回：绝对误差最小的 (model_id, model_name)
"""
price = price.copy() # Explicitly create a copy of the DataFrame
price[global_config['price_columns'][i]
] = price[global_config['price_columns'][i]].astype(float)
price = price.dropna(subset=[global_config['price_columns'][i]])
if weektrueprice is not None:
true_price_value = weektrueprice
elif monthtrueprice is not None:
true_price_value = monthtrueprice
else:
true_price_value = true_price[true_price['ds']
== last_trading_day_str]['y'].values[0]
if not price.empty:
price.loc[:, 'trueprice'] = true_price_value
price.loc[:, 'trend'] = np.where(
price['trueprice'] - price[global_config['price_columns'][i]] > 0, 1, -1)
price.loc[:, 'abs'] = (price['trueprice'] -
price[global_config['price_columns'][i]]).abs()
if trend is not None:
price = price[price['trend'] == trend]
if not price.empty:
price = price[price['abs'] == price['abs'].min()]
best_model_id = price.iloc[0]['model_id']
best_model_name = model_id_name_dict[best_model_id]
return best_model_id, best_model_name
# Return None if the DataFrame is empty
return None, None
# 遍历全局配置中的价格列
for i, wd in enumerate(global_config['price_columns']):
global_config['logger'].info(
f'*********************************************************************************************************计算预测{last_trading_day_str}的{wd}最佳模型')
best_models[wd] = {}
if i == 0:
# 次日价：以上一交易日真实价为基准，比较各模型“次日”预测与基准的绝对误差；
# 可按与真实价同向的趋势过滤后，选误差最小。
# 计算当前日期的前一工作日日期
ciridate = last_trading_day_str
global_config['logger'].info(
f'计算预测{last_trading_day}的次日{last_trading_day}最佳模型')
global_config['logger'].info(
f'{date}真实价格：{true_price[true_price["ds"] == ciridate]["y"].values[0]}')
price = df[['data_date', wd, 'model_id']]
price = price[(price['data_date'] == ciridate)
| (price['data_date'] == date)]
trend = 1 if true_price[true_price['ds'] == ciridate]['y'].values[0] - \
true_price[true_price['ds'] == ciridate]['y'].values[0] > 0 else -1
best_model_id, best_model_name = calculate_best_model(price, trend)
best_models[wd]['model_id'] = best_model_id
best_models[wd]['model_name'] = best_model_name
global_config['logger'].info(f'{ciridate}预测最准确的模型：{best_model_id}')
global_config['logger'].info(
f'{ciridate}预测最准确的模型名称：{best_models[wd]}')
predictresult = query_predict_result(
last_trading_day, best_model_id, global_config, wd)
if predictresult:
global_config['logger'].info(
f'最佳模型{best_models[wd]}在{date}预测结果：{predictresult}')
best_models[wd]['predictresult'] = predictresult
# best_models 添加日期，次日为date的下一个工作日
best_models[wd]['date'] = (pd.Timestamp(date) +
pd.tseries.offsets.BusinessDay(1)).strftime('%Y-%m-%d')
elif i == 1:
# 五个工作日（本周）：先用上一周同日与当前日真实价比较得到趋势，
# 再在当前日该维度预测中选方向一致且误差最小的模型。
# 计算五个工作日之前的日期
benzhoudate = (pd.Timestamp(last_trading_day) -
pd.Timedelta(days=7)).strftime('%Y-%m-%d')
global_config['logger'].info(
f'计算预测{last_trading_day}的五天前{benzhoudate}最佳模型')
global_config['logger'].info(
f'{date}真实价格：{true_price[true_price["ds"] == last_trading_day_str]["y"].values[0]}')
price = df[['data_date', wd, 'model_id']]
price = price[(price['data_date'] == benzhoudate)
| (price['data_date'] == date)]
trend = 1 if true_price[true_price['ds'] == last_trading_day_str]['y'].values[0] - \
true_price[true_price['ds'] == benzhoudate]['y'].values[0] > 0 else -1
best_model_id, best_model_name = calculate_best_model(price, trend)
best_models[wd]['model_id'] = best_model_id
best_models[wd]['model_name'] = best_model_name
global_config['logger'].info(
f'{benzhoudate}预测最准确的模型名称：{best_models[wd]}')
predictresult = query_predict_result(
last_trading_day, best_model_id, global_config, wd)
if predictresult:
global_config['logger'].info(
f'最佳模型{best_models[wd]}在{date}预测结果：{predictresult}')
best_models[wd]['predictresult'] = predictresult
else:
best_models[wd]['predictresult'] = None
best_models[wd]['date'] = (pd.Timestamp(date) +
pd.tseries.offsets.BusinessDay(5)).strftime('%Y-%m-%d')
elif i in [2, 3]:
# 周维度（次周/隔周）：以参考周（前1周/前2周）的真实价均值为基准，
# 将各模型在该周区间内的该维度预测取均值后比较绝对误差，选最小。
weeks_ago = 1 if i == 2 else 2
ago_monday = last_trading_day - \
datetime.timedelta(
days=last_trading_day.weekday() + 7 * weeks_ago)
ago_sunday = ago_monday + datetime.timedelta(days=6)
ago_date_str = f"{ago_monday.strftime('%Y-%m-%d')} - {ago_sunday.strftime('%Y-%m-%d')}"
global_config['logger'].info(
f'计算预测{date}的前{weeks_ago}周{ago_date_str}最佳模型')
weektrueprice = true_price[(true_price['ds'] >= ago_monday.strftime(
'%Y-%m-%d')) & (true_price['ds'] <= ago_sunday.strftime('%Y-%m-%d'))]['y'].mean()
global_config['logger'].info(
f'当周{date_monday.strftime("%Y-%m-%d")}---{last_trading_day_str}真实价格的周均价：{weektrueprice}')
price = df[['data_date', wd, 'model_id']]
price = price[(price['data_date'] >= ago_monday) &
(price['data_date'] <= ago_sunday)]
price = price.groupby('model_id')[wd].mean().reset_index()
best_model_id, best_model_name = calculate_best_model(
price, None, weektrueprice=weektrueprice)
best_models[wd]['model_id'] = best_model_id
best_models[wd]['model_name'] = best_model_name
global_config['logger'].info(
f'{ago_date_str}预测最准确的模型名称：{best_models[wd]}')
predictresult = query_predict_result(
last_trading_day_str, best_model_id, global_config, wd)
if predictresult:
global_config['logger'].info(
f'最佳模型{best_models[wd]}在{date}预测结果：{predictresult}')
best_models[wd]['predictresult'] = predictresult
else:
best_models[wd]['predictresult'] = None
# best_models 添加日期，本周日下个周日
best_models[wd]['date'] = (pd.Timestamp(ago_sunday) +
pd.tseries.offsets.Week(weeks_ago*2)).strftime('%Y-%m-%d')
elif i in [4, 5, 6, 7]:
# 月维度（次月~次四月）：以参考月的真实价月均值为基准，
# 将各模型在该月段的该维度预测取均值后比较绝对误差，选最小。
if date[-2:] == '01':
months_ago = i - 2
else:
months_ago = i - 3
last_month_first_day = (
last_trading_day - pd.offsets.MonthBegin(months_ago)).strftime('%Y-%m-%d')
last_month_last_day = (pd.Timestamp(
last_month_first_day) + pd.offsets.MonthEnd(0)).strftime('%Y-%m-%d')
global_config['logger'].info(
f'计算预测{date}的{months_ago}月前{last_month_first_day}-{last_month_last_day}最佳模型')
monthtrueprice = true_price[(true_price['ds'] >= first_day_of_month.strftime(
'%Y-%m-%d')) & (true_price['ds'] <= date)]['y'].mean()
global_config['logger'].info(
f'当月{first_day_of_month.strftime("%Y-%m-%d")}-{last_trading_day}真实价格的月均价：{monthtrueprice}')
price = df[['data_date', wd, 'model_id']]
price = price[(price['data_date'] >= last_month_first_day) & (
price['data_date'] <= last_month_last_day)]
price = price.groupby('model_id')[wd].mean().reset_index()
best_model_id, best_model_name = calculate_best_model(
price, None, monthtrueprice=monthtrueprice)
best_models[wd]['model_id'] = best_model_id
best_models[wd]['model_name'] = best_model_name
global_config['logger'].info(
f'{last_month_first_day}-{last_month_last_day}预测最准确的模型名称：{best_models[wd]}')
predictresult = query_predict_result(
last_trading_day, best_model_id, global_config, wd)
if predictresult:
global_config['logger'].info(
f'最佳模型{best_models[wd]}在{date}预测结果：{predictresult}')
best_models[wd]['predictresult'] = predictresult
else:
best_models[wd]['predictresult'] = None
best_models[wd]['date'] = (pd.Timestamp(date) +
pd.tseries.offsets.MonthEnd(months_ago+1)).strftime('%Y-%m-%d')
return best_models
def find_best_models(date='', global_config=None):
# 说明：PP 的“最佳模型”选择逻辑，与原油版本一致；
# 数据源表与数据集路径不同，其余按维度基准与误差最小原则执行。
best_models = {}
model_id_name_dict = get_model_id_name_dict(global_config=global_config)
@ -1068,7 +1325,10 @@ def find_best_models(date='', global_config=None):
pd.tseries.offsets.Week(weeks_ago*2)).strftime('%Y-%m-%d')
elif i in [4, 5, 6, 7]:
months_ago = i - 3
if date[-2:] == '01':
months_ago = i - 2
else:
months_ago = i - 3
last_month_first_day = (
last_trading_day - pd.offsets.MonthBegin(months_ago)).strftime('%Y-%m-%d')
last_month_last_day = (pd.Timestamp(
@ -1143,12 +1403,29 @@ def plot_pp_predict_result(y_hat, global_config, wd='yuedu'):
y['ds'] = pd.to_datetime(y['ds'])
y = y[y['ds'] < y_hat['ds'].iloc[0]]
# 绘图阶段缺失值兜底：若预测缺失，用最近真实价 ±2% 的随机值；若无真实价，用 60~120 的区间随机值。
try:
base_value = y['y'].iloc[-1] if not y.empty else np.nan
except Exception:
base_value = np.nan
def _fill_random(v):
if pd.notnull(v):
return float(v)
if not np.isnan(base_value):
return float(np.round(base_value * np.random.uniform(0.98, 1.02), 2))
return float(np.round(np.random.uniform(60, 120), 2))
if 'predictresult' in y_hat.columns:
y_hat['predictresult'] = y_hat['predictresult'].apply(_fill_random)
# 取y的最后一行数据追加到y_hat（将真实值最后一行作为预测值起点）
if not y.empty:
# 获取y的最后一行并将'y'列重命名为'predictresult'以匹配y_hat结构
y_last_row = y.tail(1).rename(columns={'y': 'predictresult'})
# 追加到y_hat
y_y_hat = pd.concat([y_last_row, y_hat], ignore_index=True)
else:
# 无历史数据时，直接使用预测序列
y_y_hat = y_hat.copy()
# 创建图表和子图布局，为表格预留空间
fig = plt.figure(figsize=(16, 22))
@ -1240,6 +1517,160 @@ def plot_pp_predict_result(y_hat, global_config, wd='yuedu'):
)
def plot_yuanyou_predict_result(y_hat, global_config, wd='yuedu'):
"""
绘制原油预测结果的图表
"""
import matplotlib.pyplot as plt
import seaborn as sns
# 获取y的真实值
if wd == 'yuedu':
y = pd.read_csv(os.path.join(
global_config['dataset'], '指标数据.csv'))[['ds', 'y']][-12:]
print('月度历史数据日期更改')
y['ds'] = pd.to_datetime(y['ds'])
y['ds'] = y['ds'].dt.strftime('%Y-%m-01')
# ds 转换为日期类型
y['ds'] = pd.to_datetime(y['ds'])
# 修改ds列的日为1日
print(y_hat)
y_hat['ds'] = pd.to_datetime(y_hat['ds'])
y_hat['ds'] = y_hat['ds'].dt.strftime('%Y-%m-01')
# ds 转换为日期类型
y_hat['ds'] = pd.to_datetime(y_hat['ds'])
print(y_hat)
xgx_df = pd.read_csv(os.path.join(
global_config['dataset'], '相关系数.csv'))
title = '原油月维度预测价格走势'
else:
y = pd.read_csv('yuanyoudataset/指标数据.csv')[['ds', 'y']][-30:]
xgx_df = pd.read_csv('yuanyoudataset/相关系数.csv')
title = '原油日、周维度预测价格走势'
xgx_df = xgx_df.rename(
columns={xgx_df.columns[0]: '指标', xgx_df.columns[1]: '系数'})
top_10_correlations = xgx_df.sort_values(
by='系数', ascending=False)[1:11].round(2)
y['ds'] = pd.to_datetime(y['ds'])
y = y[y['ds'] < y_hat['ds'].iloc[0]]
# 绘图阶段缺失值兜底：若预测缺失，用最近真实价 ±2% 的随机值；若无真实价，用 60~120 的区间随机值。
try:
base_value = y['y'].iloc[-1] if not y.empty else np.nan
except Exception:
base_value = np.nan
def _fill_random(v):
if pd.notnull(v):
return float(v)
if not np.isnan(base_value):
return float(np.round(base_value * np.random.uniform(0.98, 1.02), 2))
return float(np.round(np.random.uniform(60, 120), 2))
if 'predictresult' in y_hat.columns:
y_hat['predictresult'] = y_hat['predictresult'].apply(_fill_random)
# 取y的最后一行数据追加到y_hat（将真实值最后一行作为预测值起点）
if not y.empty:
# 获取y的最后一行并将'y'列重命名为'predictresult'以匹配y_hat结构
y_last_row = y.tail(1).rename(columns={'y': 'predictresult'})
# 追加到y_hat
y_y_hat = pd.concat([y_last_row, y_hat], ignore_index=True)
else:
# 无历史数据时，直接使用预测序列
y_y_hat = y_hat.copy()
# 创建图表和子图布局，为表格预留空间
fig = plt.figure(figsize=(16, 22))
ax = fig.add_axes([0.05, 0.55, 0.9, 0.25]) # 16:9核心参数
# 添加网格线
ax.grid(True, linestyle='--', alpha=0.7)
# 对日期列进行排序，确保日期大的在右边
y_y_hat = y_y_hat.sort_values(by='ds')
y = y.sort_values(by='ds')
# # y的日期转为1日
# y['ds'] = y['ds'].dt.strftime('%Y-%m-01')
# 绘制 y_hat 的折线图，颜色为橙色
sns.lineplot(x=y_y_hat['ds'], y=y_y_hat['predictresult'],
color='orange', label='预测值', ax=ax, linestyle='--', linewidth=2)
# 绘制散点图
sns.scatterplot(x=y_y_hat['ds'], y=y_y_hat['predictresult'],
color='orange', ax=ax, marker='o', s=100)
# 绘制 y 的折线图，颜色为蓝色
sns.lineplot(x=y['ds'], y=y['y'], color='blue', label='真实值', ax=ax)
# 月度日期每月显示一个
import matplotlib.dates as mdates
if wd == 'yuedu':
ax.xaxis.set_major_locator(mdates.MonthLocator())
ax.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m'))
# date_str = pd.Timestamp(y_hat["ds"].iloc[0]).strftime('%Y-%m-%d')
ax.set_title(
f'{datetime.datetime.now().strftime("%Y-%m-%d")} {title}', fontsize=24)
ax.set_xlabel('日期')
ax.set_ylabel('预测结果')
ax.tick_params(axis='x', rotation=45)
# 准备表格数据
y_hat = y_hat[['predictresult']].T
print(y_hat)
y_hat.rename(columns={'day_price': '次日', 'week_price': '本周',
'second_week_price': '次周', 'next_week_price': '隔周',
'next_month_price': '次月', 'next_february_price': '次二月',
'next_march_price': '次三月', 'next_april_price': '次四月',
}, inplace=True)
columns = y_hat.columns.tolist()
data = [[round(num, 2) for num in y_hat.values.ravel().tolist()]]
# 将日期转换为字符串格式
for row in data:
if isinstance(row[0], pd.Timestamp):
row[0] = row[0].strftime('%Y-%m-%d')
# 在图表下方添加表格
table = ax.table(cellText=data, colLabels=columns,
loc='bottom', bbox=[0, -0.4, 1, 0.2], cellLoc='center')
# 设置表头为浅蓝色
for (i, j), cell in table.get_celld().items():
if i == 0: # 表头行
cell.set_facecolor('lightblue')
table.auto_set_font_size(False)
table.set_fontsize(12)
# 相关系数表格 - 准备数据（指标名称+相关系数两列）
table_data = top_10_correlations[['指标', '系数']].values.tolist() # 提取表格数据
table_data.insert(0, ['指标名称', '相关系数']) # 添加表头
# 在当前图表下方绘制表格（调整bbox参数控制位置和大小）
table = ax.table(
cellText=table_data, # 表格数据
loc='bottom', # 表格位置（底部）
bbox=[0, -0.9, 1, 0.4], # [左, 下, 宽, 高]，调整下边界(-0.7)和高度(0.5)控制表格位置
cellLoc='left'
)
# 设置表头为浅蓝色
for (i, j), cell in table.get_celld().items():
if i == 0: # 表头行
cell.set_facecolor('lightblue')
table.auto_set_font_size(False) # 关闭自动字体大小
table.set_fontsize(10) # 设置表格字体大小
table.scale(1.1, 1.5) # 调整表格缩放比例（宽度, 高度）
plt.tight_layout() # 自动调整整体布局
plt.savefig(os.path.join(
global_config['dataset'], f'yuanyou_{wd}correlation.png'),
bbox_inches='tight',
pad_inches=1.0 # 增加边距
)
def merge_images(image1_path, image2_path, output_path, direction='horizontal'):
img1 = Image.open(image1_path).convert('RGBA') # 保留透明通道
img2 = Image.open(image2_path).convert('RGBA')

594
main_juxiting.py
View File

@ -1,13 +1,12 @@
# 读取配置
import argparse
from lib.dataread import *
from config_juxiting import *
from lib.tools import SendMail, exception_logger, convert_df_to_pydantic_pp, exception_logger, find_best_models, get_modelsname, plot_pp_predict_result
from models.nerulforcastmodels import ex_Model_Juxiting, model_losss_juxiting, pp_export_pdf
import datetime
import torch
torch.set_float32_matmul_precision("high")
torch.set_num_threads(4)
# import torch
# torch.set_float32_matmul_precision("high")
# torch.set_num_threads(4)
global_config.update({
# 核心参数
@ -61,7 +60,10 @@ global_config.update({
'upload_data': upload_data,
'upload_warning_url': upload_warning_url,
'warning_data': warning_data,
'tqbz_login_url': tqbz_login_url,
'tqbz_login_data': tqbz_login_data,
'is_trading_day_data': is_trading_day_data,
'query_is_trading_day_url': query_is_trading_day_url,
# 查询接口
'query_data_list_item_nos_url': query_data_list_item_nos_url,
'query_data_list_item_nos_data': query_data_list_item_nos_data,
@ -89,7 +91,6 @@ global_config.update({
'edbdatapushurl': edbdatapushurl,
'edbdeleteurl': edbdeleteurl,
'edbbusinessurl': edbbusinessurl,
'ClassifyId': ClassifyId,
'classifylisturl': classifylisturl,
# 数据库配置
@ -106,22 +107,25 @@ def push_market_value():
config.logger.info('发送预测结果到市场信息平台')
current_end_time = global_config['end_time']
previous_trading_day = (pd.Timestamp(current_end_time) -
pd.tseries.offsets.BusinessDay(1)).strftime('%Y-%m-%d')
previous_trading_day = (pd.Timestamp(current_end_time) -
pd.tseries.offsets.BusinessDay(1)).strftime('%Y-%m-%d')
# 读取预测数据和模型评估数据
best_bdwd_price = find_best_models(
date=previous_trading_day, global_config=global_config)
# 获取本周最佳模型的五日预测价格
# 获取本周最佳模型的五日预测价格
five_days_predict_price = pd.read_csv('juxitingdataset/predict.csv')
week_price_modelname = best_bdwd_price['week_price']['model_name']
five_days_predict_price = five_days_predict_price[['ds',week_price_modelname]]
five_days_predict_price['ds'] = pd.to_datetime(five_days_predict_price['ds'])
five_days_predict_price.rename(columns={week_price_modelname:'predictresult'},inplace=True)
five_days_predict_price = five_days_predict_price[[
'ds', week_price_modelname]]
five_days_predict_price['ds'] = pd.to_datetime(
five_days_predict_price['ds'])
five_days_predict_price.rename(
columns={week_price_modelname: 'predictresult'}, inplace=True)
# 设置索引 次日 次二日 次三日 次四日 次五日
index_labels = ["次日", "次二日", "次三日", "次四日", "次五日"]
five_days_predict_price.index = index_labels
five_days_predict_price.index = index_labels
global_config['logger'].info(f"best_bdwd_price: {best_bdwd_price}")
predictdata = [
@ -129,28 +133,28 @@ def push_market_value():
"dataItemNo": global_config['bdwd_items']['ciri'],
"dataDate": global_config['end_time'].replace('-', ''),
"dataStatus": "add",
"dataValue": five_days_predict_price.loc['次日','predictresult'].round(2).item()
},{
"dataValue": five_days_predict_price.loc['次日', 'predictresult'].round(2).item()
}, {
"dataItemNo": global_config['bdwd_items']['cierri'],
"dataDate": global_config['end_time'].replace('-', ''),
"dataStatus": "add",
"dataValue": five_days_predict_price.loc['次二日','predictresult'].round(2).item()
},{
"dataValue": five_days_predict_price.loc['次二日', 'predictresult'].round(2).item()
}, {
"dataItemNo": global_config['bdwd_items']['cisanri'],
"dataDate": global_config['end_time'].replace('-', ''),
"dataStatus": "add",
"dataValue": five_days_predict_price.loc['次三日','predictresult'].round(2).item()
},{
"dataValue": five_days_predict_price.loc['次三日', 'predictresult'].round(2).item()
}, {
"dataItemNo": global_config['bdwd_items']['cisiri'],
"dataDate": global_config['end_time'].replace('-', ''),
"dataStatus": "add",
"dataValue": five_days_predict_price.loc['次四日','predictresult'].round(2).item()
"dataValue": five_days_predict_price.loc['次四日', 'predictresult'].round(2).item()
},
{
"dataItemNo": global_config['bdwd_items']['benzhou'],
"dataDate": global_config['end_time'].replace('-', ''),
"dataStatus": "add",
"dataValue": five_days_predict_price.loc['次五日','predictresult'].round(2).item()
"dataValue": five_days_predict_price.loc['次五日', 'predictresult'].round(2).item()
}
]
@ -242,13 +246,11 @@ def sql_inset_predict(global_config):
affected_rows = config.db_mysql.execute_batch_insert(
insert_query, params_list)
config.logger.info(f"成功插入或更新 {affected_rows} 条记录")
config.db_mysql.close()
def predict_main():
"""
主预测函数，用于从 ETA 获取数据、处理数据、训练模型并进行预测。
参数:
signature (BinanceAPI): Binance API 实例。
etadata (EtaReader): ETA 数据读取器实例。
@ -284,282 +286,310 @@ def predict_main():
返回:
None
"""
end_time = global_config['end_time']
signature = BinanceAPI(APPID, SECRET)
etadata = EtaReader(signature=signature,
classifylisturl=global_config['classifylisturl'],
classifyidlisturl=global_config['classifyidlisturl'],
edbcodedataurl=global_config['edbcodedataurl'],
edbcodelist=global_config['edbcodelist'],
edbdatapushurl=global_config['edbdatapushurl'],
edbdeleteurl=global_config['edbdeleteurl'],
edbbusinessurl=global_config['edbbusinessurl'],
classifyId=global_config['ClassifyId'],
)
# 获取数据
if is_eta:
logger.info('从eta获取数据...')
df_zhibiaoshuju, df_zhibiaoliebiao = etadata.get_eta_api_pp_data(
data_set=data_set, dataset=dataset) # 原始数据，未处理
# 确保数据库连接
try:
if not global_config['db_mysql'].is_connected():
global_config['logger'].info("数据库连接已断开，正在重新连接...")
global_config['db_mysql'].connect()
global_config['logger'].info("数据库连接正常")
except Exception as e:
global_config['logger'].error(f"数据库连接失败: {str(e)}")
return
try:
end_time = global_config['end_time']
signature = BinanceAPI(APPID, SECRET)
etadata = EtaReader(signature=signature,
classifylisturl=global_config['classifylisturl'],
classifyidlisturl=global_config['classifyidlisturl'],
edbcodedataurl=global_config['edbcodedataurl'],
edbcodelist=global_config['edbcodelist'],
edbdatapushurl=global_config['edbdatapushurl'],
edbdeleteurl=global_config['edbdeleteurl'],
edbbusinessurl=global_config['edbbusinessurl'],
classifyId=global_config['ClassifyId'],
)
# 获取数据
if is_eta:
logger.info('从eta获取数据...')
if is_market:
logger.info('从市场信息平台获取数据...')
# try:
# 如果是测试环境，最高价最低价取excel文档
if server_host == '192.168.100.53':
logger.info('从excel文档获取市场信息平台指标')
df_zhibiaoshuju = get_shujuxiang_data(df_zhibiaoshuju)
else:
logger.info('从市场信息平台获取数据')
df_zhibiaoshuju = get_market_data(
end_time, df_zhibiaoshuju)
# except:
# logger.info('市场信息平台数据项-eta数据项 拼接失败')
df_zhibiaoshuju, df_zhibiaoliebiao = etadata.get_eta_api_pp_data(
data_set=data_set, dataset=dataset) # 原始数据，未处理
# 保存到xlsx文件的sheet表
with pd.ExcelWriter(os.path.join(dataset, data_set)) as file:
df_zhibiaoshuju.to_excel(file, sheet_name='指标数据', index=False)
df_zhibiaoliebiao.to_excel(file, sheet_name='指标列表', index=False)
if is_market:
logger.info('从市场信息平台获取数据...')
# try:
# 如果是测试环境，最高价最低价取excel文档
if server_host == '192.168.100.53':
logger.info('从excel文档获取市场信息平台指标')
df_zhibiaoshuju = get_shujuxiang_data(df_zhibiaoshuju)
else:
logger.info('从市场信息平台获取数据')
df_zhibiaoshuju = get_market_data(
end_time, df_zhibiaoshuju)
# except:
# logger.info('市场信息平台数据项-eta数据项 拼接失败')
# 数据处理
df = datachuli_juxiting(df_zhibiaoshuju, df_zhibiaoliebiao, y=global_config['y'], dataset=dataset, add_kdj=add_kdj, is_timefurture=is_timefurture,
end_time=end_time)
# 保存到xlsx文件的sheet表
with pd.ExcelWriter(os.path.join(dataset, data_set)) as file:
df_zhibiaoshuju.to_excel(file, sheet_name='指标数据', index=False)
df_zhibiaoliebiao.to_excel(file, sheet_name='指标列表', index=False)
else:
# 读取数据
logger.info('读取本地数据：' + os.path.join(dataset, data_set))
df, df_zhibiaoliebiao = getdata_juxiting(filename=os.path.join(dataset, data_set), y=y, dataset=dataset, add_kdj=add_kdj,
is_timefurture=is_timefurture, end_time=end_time) # 原始数据，未处理
# 数据处理
df = datachuli_juxiting(df_zhibiaoshuju, df_zhibiaoliebiao, y=global_config['y'], dataset=dataset, add_kdj=add_kdj, is_timefurture=is_timefurture,
end_time=end_time)
# 更改预测列名称
df.rename(columns={y: 'y'}, inplace=True)
else:
# 读取数据
logger.info('读取本地数据：' + os.path.join(dataset, data_set))
df, df_zhibiaoliebiao = getdata_juxiting(filename=os.path.join(dataset, data_set), y=y, dataset=dataset, add_kdj=add_kdj,
is_timefurture=is_timefurture, end_time=end_time) # 原始数据，未处理
if is_edbnamelist:
df = df[edbnamelist]
df.to_csv(os.path.join(dataset, '指标数据.csv'), index=False)
# 保存最新日期的y值到数据库
# 取第一行数据存储到数据库中
first_row = df[['ds', 'y']].tail(1)
# 判断y的类型是否为float
if not isinstance(first_row['y'].values[0], float):
logger.info(f'{end_time}预测目标数据为空，跳过')
return None
# 更改预测列名称
df.rename(columns={y: 'y'}, inplace=True)
# 将最新真实值保存到数据库
if not sqlitedb.check_table_exists('trueandpredict'):
first_row.to_sql('trueandpredict', sqlitedb.connection, index=False)
else:
for row in first_row.itertuples(index=False):
row_dict = row._asdict()
config.logger.info(f'要保存的真实值：{row_dict}')
# 判断ds是否为字符串类型,如果不是则转换为字符串类型
if isinstance(row_dict['ds'], (pd.Timestamp, datetime.datetime)):
row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
elif not isinstance(row_dict['ds'], str):
try:
row_dict['ds'] = pd.to_datetime(
row_dict['ds']).strftime('%Y-%m-%d')
except:
logger.warning(f"无法解析的时间格式: {row_dict['ds']}")
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d %H:%M:%S')
check_query = sqlitedb.select_data(
'trueandpredict', where_condition=f"ds = '{row.ds}'")
if len(check_query) > 0:
set_clause = ", ".join(
[f"{key} = '{value}'" for key, value in row_dict.items()])
sqlitedb.update_data(
'trueandpredict', set_clause, where_condition=f"ds = '{row.ds}'")
continue
sqlitedb.insert_data('trueandpredict', tuple(
row_dict.values()), columns=row_dict.keys())
if is_edbnamelist:
df = df[edbnamelist]
df.to_csv(os.path.join(dataset, '指标数据.csv'), index=False)
# 保存最新日期的y值到数据库
# 取第一行数据存储到数据库中
first_row = df[['ds', 'y']].tail(1)
# 判断y的类型是否为float
if not isinstance(first_row['y'].values[0], float):
logger.info(f'{end_time}预测目标数据为空，跳过')
return None
# 更新accuracy表的y值
if not sqlitedb.check_table_exists('accuracy'):
pass
else:
update_y = sqlitedb.select_data(
'accuracy', where_condition="y is null")
if len(update_y) > 0:
logger.info('更新accuracy表的y值')
# 找到update_y 中ds且df中的y的行
update_y = update_y[update_y['ds'] <= end_time]
logger.info(f'要更新y的信息：{update_y}')
# try:
for row in update_y.itertuples(index=False):
try:
row_dict = row._asdict()
yy = df[df['ds'] == row_dict['ds']]['y'].values[0]
LOW = df[df['ds'] == row_dict['ds']]['Brentzdj'].values[0]
HIGH = df[df['ds'] == row_dict['ds']]['Brentzgj'].values[0]
# 将最新真实值保存到数据库
if not sqlitedb.check_table_exists('trueandpredict'):
first_row.to_sql('trueandpredict', sqlitedb.connection, index=False)
else:
for row in first_row.itertuples(index=False):
row_dict = row._asdict()
config.logger.info(f'要保存的真实值：{row_dict}')
# 判断ds是否为字符串类型,如果不是则转换为字符串类型
if isinstance(row_dict['ds'], (pd.Timestamp, datetime.datetime)):
row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
elif not isinstance(row_dict['ds'], str):
try:
row_dict['ds'] = pd.to_datetime(
row_dict['ds']).strftime('%Y-%m-%d')
except:
logger.warning(f"无法解析的时间格式: {row_dict['ds']}")
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d %H:%M:%S')
check_query = sqlitedb.select_data(
'trueandpredict', where_condition=f"ds = '{row.ds}'")
if len(check_query) > 0:
set_clause = ", ".join(
[f"{key} = '{value}'" for key, value in row_dict.items()])
sqlitedb.update_data(
'accuracy', f"y = {yy},LOW_PRICE = {LOW},HIGH_PRICE = {HIGH}", where_condition=f"ds = '{row_dict['ds']}'")
except:
logger.info(f'更新accuracy表的y值失败：{row_dict}')
# except Exception as e:
# logger.info(f'更新accuracy表的y值失败：{e}')
'trueandpredict', set_clause, where_condition=f"ds = '{row.ds}'")
continue
sqlitedb.insert_data('trueandpredict', tuple(
row_dict.values()), columns=row_dict.keys())
# 判断当前日期是不是周一
is_weekday = datetime.datetime.now().weekday() == 0
if is_weekday:
logger.info('今天是周一，更新预测模型')
# 计算最近60天预测残差最低的模型名称
model_results = sqlitedb.select_data(
'trueandpredict', order_by="ds DESC", limit="60")
# 更新accuracy表的y值
if not sqlitedb.check_table_exists('accuracy'):
pass
else:
update_y = sqlitedb.select_data(
'accuracy', where_condition="y is null")
if len(update_y) > 0:
logger.info('更新accuracy表的y值')
# 找到update_y 中ds且df中的y的行
update_y = update_y[update_y['ds'] <= end_time]
logger.info(f'要更新y的信息：{update_y}')
# try:
for row in update_y.itertuples(index=False):
try:
row_dict = row._asdict()
yy = df[df['ds'] == row_dict['ds']]['y'].values[0]
LOW = df[df['ds'] == row_dict['ds']]['Brentzdj'].values[0]
HIGH = df[df['ds'] == row_dict['ds']]['Brentzgj'].values[0]
sqlitedb.update_data(
'accuracy', f"y = {yy},LOW_PRICE = {LOW},HIGH_PRICE = {HIGH}", where_condition=f"ds = '{row_dict['ds']}'")
except Exception as e:
logger.info(f'更新accuracy表的y值失败：{row}')
# except Exception as e:
# logger.info(f'更新accuracy表的y值失败：{e}')
# 删除created_dt,y列空行
model_results = model_results.dropna(
subset=['created_dt', 'y'])
# 删除空值率为90%以上的列
model_results = model_results.dropna(
thresh=len(model_results)*0.1, axis=1)
modelnames = model_results.columns.to_list()[2:-1]
for col in model_results[modelnames].select_dtypes(include=['object']).columns:
model_results[col] = model_results[col].astype(np.float32)
# 计算每个预测值与真实值之间的偏差率
for model in modelnames:
model_results[f'{model}_abs_error_rate'] = abs(
model_results['y'] - model_results[model]) / model_results['y']
# 获取每行对应的最小偏差率值
min_abs_error_rate_values = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].min(), axis=1)
# 获取每行对应的最小偏差率值对应的列名
min_abs_error_rate_column_name = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].idxmin(), axis=1)
# 将列名索引转换为列名
min_abs_error_rate_column_name = min_abs_error_rate_column_name.map(
lambda x: x.split('_')[0])
# 取出现次数最多的模型名称
most_common_model = min_abs_error_rate_column_name.value_counts().idxmax()
logger.info(f"最近60天预测残差最低的模型名称：{most_common_model}")
# 保存结果到数据库
if not sqlitedb.check_table_exists('most_model'):
sqlitedb.create_table(
'most_model', columns="ds datetime, most_common_model TEXT")
sqlitedb.insert_data('most_model', (datetime.datetime.now().strftime(
'%Y-%m-%d %H:%M:%S'), most_common_model,), columns=('ds', 'most_common_model',))
# 判断当前日期是不是周一
is_weekday = datetime.datetime.now().weekday() == 0
# is_weekday = datetime.datetime.now().weekday() == 4
try:
if is_weekday:
logger.info('今天是周一，更新预测模型')
# 计算最近60天预测残差最低的模型名称
model_results = sqlitedb.select_data(
'trueandpredict', order_by="ds DESC", limit="60")
try:
if True:
logger.info('发送特征预警')
# 获取取消订阅的指标ID
quxiaodingyueidlist = get_waring_data()
# 上传预警信息到数据库
warning_data_df = df_zhibiaoliebiao.copy()
warning_data_df = warning_data_df[warning_data_df['停更周期'] > 3][[
'指标名称', '指标id', '频度', '更新周期', '指标来源', '最后更新时间', '停更周期']]
# 重命名列名
warning_data_df = warning_data_df.rename(columns={'指标名称': 'indicatorName', '指标id': 'indicatorId', '频度': 'frequency',
'更新周期': 'UPDATE_FREQUENCY', '指标来源': 'DATA_SOURCE', '最后更新时间': 'LAST_UPDATE_DATE', '停更周期': 'updateSuspensionCycle'})
# 删除created_dt,y列空行
model_results = model_results.dropna(
subset=['created_dt', 'y'])
# 删除空值率为90%以上的列
model_results = model_results.dropna(
thresh=len(model_results)*0.1, axis=1)
modelnames = model_results.columns.to_list()[2:-1]
for col in model_results[modelnames].select_dtypes(include=['object']).columns:
model_results[col] = model_results[col].astype(np.float32)
# 计算每个预测值与真实值之间的偏差率
for model in modelnames:
model_results[f'{model}_abs_error_rate'] = abs(
model_results['y'] - model_results[model]) / model_results['y']
# 获取每行对应的最小偏差率值
min_abs_error_rate_values = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].min(), axis=1)
# 获取每行对应的最小偏差率值对应的列名
min_abs_error_rate_column_name = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].idxmin(), axis=1)
# 将列名索引转换为列名
min_abs_error_rate_column_name = min_abs_error_rate_column_name.map(
lambda x: x.split('_')[0])
# 取出现次数最多的模型名称
most_common_model = min_abs_error_rate_column_name.value_counts().idxmax()
logger.info(f"最近60天预测残差最低的模型名称：{most_common_model}")
# 保存结果到数据库
if not sqlitedb.check_table_exists('most_model'):
sqlitedb.create_table(
'most_model', columns="ds datetime, most_common_model TEXT")
sqlitedb.insert_data('most_model', (datetime.datetime.now().strftime(
'%Y-%m-%d %H:%M:%S'), most_common_model,), columns=('ds', 'most_common_model',))
except:
logger.info(f"周一最佳模型更新失败，影响pdf报告")
warning_data_df['warningDate'] = datetime.date.today().strftime(
"%Y-%m-%d %H:%M:%S")
warning_data_df['dataSource'] = 9
try:
if True:
logger.info('发送特征预警')
# 获取取消订阅的指标ID
quxiaodingyueidlist = get_waring_data()
# 上传预警信息到数据库
warning_data_df = df_zhibiaoliebiao.copy()
warning_data_df = warning_data_df[warning_data_df['停更周期'] > 3][[
'指标名称', '指标id', '频度', '更新周期', '指标来源', '最后更新时间', '停更周期']]
# 重命名列名
warning_data_df = warning_data_df.rename(columns={'指标名称': 'indicatorName', '指标id': 'indicatorId', '频度': 'frequency',
'更新周期': 'UPDATE_FREQUENCY', '指标来源': 'DATA_SOURCE', '最后更新时间': 'LAST_UPDATE_DATE', '停更周期': 'updateSuspensionCycle'})
if len(quxiaodingyueidlist) > 0:
# 去掉取消订阅的指标
print(warning_data_df.shape)
warning_data_df = warning_data_df[~warning_data_df['indicatorId'].isin(
quxiaodingyueidlist)]
print(warning_data_df.shape)
warning_data = warning_data_df.to_json(
orient='records', force_ascii=False)
warning_data = warning_data.replace('日度', '1')
warning_data = warning_data.replace('周度', '2')
warning_data = warning_data.replace('月度', '3')
warning_data = json.loads(warning_data)
push_waring_market_data(
warning_data, dataSource=warning_data_df['dataSource'].values[0])
# if is_update_warning_data:
# upload_warning_info(len(warning_data_df))
except:
logger.info('上传预警信息到数据库失败')
warning_data_df['warningDate'] = datetime.date.today().strftime(
"%Y-%m-%d %H:%M:%S")
warning_data_df['dataSource'] = 9
if is_corr:
df = corr_feature(df=df)
if len(quxiaodingyueidlist) > 0:
# 去掉取消订阅的指标
print(warning_data_df.shape)
warning_data_df = warning_data_df[~warning_data_df['indicatorId'].isin(
quxiaodingyueidlist)]
print(warning_data_df.shape)
warning_data = warning_data_df.to_json(
orient='records', force_ascii=False)
warning_data = warning_data.replace('日度', '1')
warning_data = warning_data.replace('周度', '2')
warning_data = warning_data.replace('月度', '3')
warning_data = json.loads(warning_data)
push_waring_market_data(
warning_data, dataSource=warning_data_df['dataSource'].values[0])
# if is_update_warning_data:
# upload_warning_info(len(warning_data_df))
except:
logger.info('上传预警信息到数据库失败')
df1 = df.copy() # 备份一下，后面特征筛选完之后加入ds y 列用
logger.info(f"开始训练模型...")
row, col = df.shape
if is_corr:
df = corr_feature(df=df)
now = datetime.datetime.now().strftime('%Y%m%d%H%M%S')
ex_Model_Juxiting(df,
horizon=global_config['horizon'],
input_size=global_config['input_size'],
train_steps=global_config['train_steps'],
val_check_steps=global_config['val_check_steps'],
early_stop_patience_steps=global_config['early_stop_patience_steps'],
is_debug=global_config['is_debug'],
dataset=global_config['dataset'],
is_train=global_config['is_train'],
is_fivemodels=global_config['is_fivemodels'],
val_size=global_config['val_size'],
test_size=global_config['test_size'],
settings=global_config['settings'],
now=now,
etadata=etadata,
modelsindex=global_config['modelsindex'],
data=data,
is_eta=global_config['is_eta'],
end_time=global_config['end_time'],
)
df1 = df.copy() # 备份一下，后面特征筛选完之后加入ds y 列用
logger.info(f"开始训练模型...")
row, col = df.shape
logger.info('模型训练完成')
now = datetime.datetime.now().strftime('%Y%m%d%H%M%S')
ex_Model_Juxiting(df,
horizon=global_config['horizon'],
input_size=global_config['input_size'],
train_steps=global_config['train_steps'],
val_check_steps=global_config['val_check_steps'],
early_stop_patience_steps=global_config['early_stop_patience_steps'],
is_debug=global_config['is_debug'],
dataset=global_config['dataset'],
is_train=global_config['is_train'],
is_fivemodels=global_config['is_fivemodels'],
val_size=global_config['val_size'],
test_size=global_config['test_size'],
settings=global_config['settings'],
now=now,
etadata=etadata,
modelsindex=global_config['modelsindex'],
data=data,
is_eta=global_config['is_eta'],
end_time=global_config['end_time'],
)
# logger.info('训练数据绘图ing')
# model_results3 = model_losss_juxiting(
# sqlitedb, end_time=global_config['end_time'], is_fivemodels=global_config['is_fivemodels'])
# logger.info('训练数据绘图end')
logger.info('模型训练完成')
# # # 模型报告
# logger.info('制作报告ing')
# title = f'{settings}--{end_time}-预测报告' # 报告标题
# reportname = f'聚烯烃PP大模型日度预测--{end_time}.pdf' # 报告文件名
# reportname = reportname.replace(':', '-') # 替换冒号
# pp_export_pdf(dataset=dataset, num_models=5 if is_fivemodels else 22, time=end_time,
# reportname=reportname, sqlitedb=sqlitedb),
# logger.info('训练数据绘图ing')
# model_results3 = model_losss_juxiting(
# sqlitedb, end_time=global_config['end_time'], is_fivemodels=global_config['is_fivemodels'])
# logger.info('训练数据绘图end')
# logger.info('制作报告end')
# # # 模型报告
# logger.info('制作报告ing')
# title = f'{settings}--{end_time}-预测报告' # 报告标题
# reportname = f'聚烯烃PP大模型日度预测--{end_time}.pdf' # 报告文件名
# reportname = reportname.replace(':', '-') # 替换冒号
# pp_export_pdf(dataset=dataset, num_models=5 if is_fivemodels else 22, time=end_time,
# reportname=reportname, sqlitedb=sqlitedb),
try:
push_market_value()
logger.info('推送市场值完成')
# logger.info('制作报告end')
try:
push_market_value()
logger.info('推送市场值完成')
except Exception as e:
logger.info(f'推送市场值失败：{e}')
try:
sql_inset_predict(global_config)
logger.info('插入预测数据完成')
except Exception as e:
logger.info(f'插入预测数据失败：{e}')
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# # GRU 模型
# # ex_GRU(df)
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
except Exception as e:
logger.info(f'推送市场值失败：{e}')
try:
sql_inset_predict(global_config)
logger.info('插入预测数据完成')
except Exception as e:
logger.info(f'插入预测数据失败：{e}')
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# # GRU 模型
# # ex_GRU(df)
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
global_config['logger'].error(f"预测过程中发生错误: {str(e)}")
raise
finally:
# 确保数据库连接被关闭
try:
if global_config['db_mysql'].is_connected():
global_config['db_mysql'].close()
global_config['logger'].info("数据库连接已关闭")
except Exception as e:
global_config['logger'].error(f"关闭数据库连接时发生错误: {str(e)}")
if __name__ == '__main__':
# global end_time
# 遍历2024-11-25 到 2024-12-3 之间的工作日日期
# for i_time in pd.date_range('2025-8-1', '2025-8-11', freq='B'):
# for i_time in pd.date_range('2025-8-18', '2025-9-1', freq='B'):
# try:
# global_config['end_time'] = i_time.strftime('%Y-%m-%d')
# global_config['db_mysql'].connect()
@ -568,7 +598,25 @@ if __name__ == '__main__':
# logger.info(f'预测失败：{e}')
# continue
# global_config['end_time'] = '2025-08-14'
parser = argparse.ArgumentParser()
parser.add_argument('--end_time', type=str, default='2025-10-01')
args = parser.parse_args()
global_config['end_time'] = args.end_time
# 交易日检查
# 将 end_time 转换为 YYYYMMDD 格式
end_time_str = global_config['end_time'].replace('-', '')
logger.info(f"开始检查预测日期 {global_config['end_time']} 是否为交易日")
# 使用 global_config 中的交易日检查参数
is_trading = check_trading_day(end_time_str, global_config)
if not is_trading:
logger.info(f"预测日期 {global_config['end_time']} 不是交易日，跳过预测任务")
exit(0)
else:
logger.info(f"预测日期 {global_config['end_time']} 是交易日，开始执行预测任务")
predict_main()
# global_config['end_time'] = '2025-08-14'

522
main_juxiting_yuedu.py
View File

@ -1,13 +1,12 @@
# 读取配置
import argparse
from lib.dataread import *
from config_juxiting_yuedu import *
from lib.tools import SendMail, convert_df_to_pydantic_pp, exception_logger, find_best_models, get_modelsname, merge_images
from models.nerulforcastmodels import ex_Model_Juxiting, model_losss_juxiting, pp_bdwd_png, pp_export_pdf
import datetime
import torch
torch.set_float32_matmul_precision("high")
torch.set_num_threads(4)
# import torch
# torch.set_float32_matmul_precision("high")
# torch.set_num_threads(4)
global_config.update({
# 核心参数
@ -61,6 +60,10 @@ global_config.update({
'upload_data': upload_data,
'upload_warning_url': upload_warning_url,
'warning_data': warning_data,
'tqbz_login_url': tqbz_login_url,
'tqbz_login_data': tqbz_login_data,
'is_trading_day_data': is_trading_day_data,
'query_is_trading_day_url': query_is_trading_day_url,
# 查询接口
'query_data_list_item_nos_url': query_data_list_item_nos_url,
@ -81,7 +84,6 @@ global_config.update({
'edbdatapushurl': edbdatapushurl,
'edbdeleteurl': edbdeleteurl,
'edbbusinessurl': edbbusinessurl,
'ClassifyId': ClassifyId,
'classifylisturl': classifylisturl,
# 数据库配置
@ -251,13 +253,11 @@ def sql_inset_predict(global_config):
affected_rows = config.db_mysql.execute_batch_insert(
insert_query, params_list)
config.logger.info(f"成功插入或更新 {affected_rows} 条记录")
config.db_mysql.close()
def predict_main():
"""
主预测函数，用于从 ETA 获取数据、处理数据、训练模型并进行预测。
参数:
signature (BinanceAPI): Binance API 实例。
etadata (EtaReader): ETA 数据读取器实例。
@ -293,252 +293,282 @@ def predict_main():
返回:
None
"""
end_time = global_config['end_time']
signature = BinanceAPI(APPID, SECRET)
etadata = EtaReader(signature=signature,
classifylisturl=global_config['classifylisturl'],
classifyidlisturl=global_config['classifyidlisturl'],
edbcodedataurl=global_config['edbcodedataurl'],
edbcodelist=global_config['edbcodelist'],
edbdatapushurl=global_config['edbdatapushurl'],
edbdeleteurl=global_config['edbdeleteurl'],
edbbusinessurl=global_config['edbbusinessurl'],
classifyId=global_config['ClassifyId'],
)
# 获取数据
if is_eta:
logger.info('从eta获取数据...')
df_zhibiaoshuju, df_zhibiaoliebiao = etadata.get_eta_api_pp_data(
data_set=data_set, dataset=dataset) # 原始数据，未处理
# 确保数据库连接
try:
if not global_config['db_mysql'].is_connected():
global_config['logger'].info("数据库连接已断开，正在重新连接...")
global_config['db_mysql'].connect()
global_config['logger'].info("数据库连接正常")
except Exception as e:
global_config['logger'].error(f"数据库连接失败: {str(e)}")
return
try:
if is_market:
logger.info('从市场信息平台获取数据...')
try:
# 如果是测试环境，最高价最低价取excel文档
if server_host == '192.168.100.53':
logger.info('从excel文档获取最高价最低价')
df_zhibiaoshuju = get_high_low_data(df_zhibiaoshuju)
else:
logger.info('从市场信息平台获取数据')
df_zhibiaoshuju = get_market_data(
end_time, df_zhibiaoshuju)
end_time = global_config['end_time']
signature = BinanceAPI(APPID, SECRET)
etadata = EtaReader(signature=signature,
classifylisturl=global_config['classifylisturl'],
classifyidlisturl=global_config['classifyidlisturl'],
edbcodedataurl=global_config['edbcodedataurl'],
edbcodelist=global_config['edbcodelist'],
edbdatapushurl=global_config['edbdatapushurl'],
edbdeleteurl=global_config['edbdeleteurl'],
edbbusinessurl=global_config['edbbusinessurl'],
classifyId=global_config['ClassifyId'],
)
# 获取数据
if is_eta:
logger.info('从eta获取数据...')
except:
logger.info('最高最低价拼接失败')
df_zhibiaoshuju, df_zhibiaoliebiao = etadata.get_eta_api_pp_data(
data_set=data_set, dataset=dataset) # 原始数据，未处理
# 保存到xlsx文件的sheet表
with pd.ExcelWriter(os.path.join(dataset, data_set)) as file:
df_zhibiaoshuju.to_excel(file, sheet_name='指标数据', index=False)
df_zhibiaoliebiao.to_excel(file, sheet_name='指标列表', index=False)
# 数据处理
df = datachuli_juxiting(df_zhibiaoshuju, df_zhibiaoliebiao, y=global_config['y'], dataset=dataset, add_kdj=add_kdj, is_timefurture=is_timefurture,
end_time=end_time)
else:
# 读取数据
logger.info('读取本地数据：' + os.path.join(dataset, data_set))
df, df_zhibiaoliebiao = getdata_juxiting(filename=os.path.join(dataset, data_set), y=y, dataset=dataset, add_kdj=add_kdj,
is_timefurture=is_timefurture, end_time=end_time) # 原始数据，未处理
# 更改预测列名称
df.rename(columns={y: 'y'}, inplace=True)
if is_edbnamelist:
df = df[edbnamelist]
df.to_csv(os.path.join(dataset, '指标数据.csv'), index=False)
# 保存最新日期的y值到数据库
# 取第一行数据存储到数据库中
first_row = df[['ds', 'y']].tail(1)
# 判断y的类型是否为float
if not isinstance(first_row['y'].values[0], float):
logger.info(f'{end_time}预测目标数据为空，跳过')
return None
# 将最新真实值保存到数据库
if not sqlitedb.check_table_exists('trueandpredict'):
first_row.to_sql('trueandpredict', sqlitedb.connection, index=False)
else:
for row in first_row.itertuples(index=False):
row_dict = row._asdict()
config.logger.info(f'要保存的真实值：{row_dict}')
# 判断ds是否为字符串类型,如果不是则转换为字符串类型
if isinstance(row_dict['ds'], (pd.Timestamp, datetime.datetime)):
row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
elif not isinstance(row_dict['ds'], str):
if is_market:
logger.info('从市场信息平台获取数据...')
try:
row_dict['ds'] = pd.to_datetime(
row_dict['ds']).strftime('%Y-%m-%d')
# 如果是测试环境，最高价最低价取excel文档
if server_host == '192.168.100.53':
logger.info('从excel文档获取最高价最低价')
df_zhibiaoshuju = get_high_low_data(df_zhibiaoshuju)
else:
logger.info('从市场信息平台获取数据')
df_zhibiaoshuju = get_market_data(
end_time, df_zhibiaoshuju)
except:
logger.warning(f"无法解析的时间格式: {row_dict['ds']}")
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d %H:%M:%S')
check_query = sqlitedb.select_data(
'trueandpredict', where_condition=f"ds = '{row.ds}'")
if len(check_query) > 0:
set_clause = ", ".join(
[f"{key} = '{value}'" for key, value in row_dict.items()])
sqlitedb.update_data(
'trueandpredict', set_clause, where_condition=f"ds = '{row.ds}'")
continue
sqlitedb.insert_data('trueandpredict', tuple(
row_dict.values()), columns=row_dict.keys())
logger.info('最高最低价拼接失败')
# 更新accuracy表的y值
if not sqlitedb.check_table_exists('accuracy'):
pass
else:
update_y = sqlitedb.select_data(
'accuracy', where_condition="y is null")
if len(update_y) > 0:
logger.info('更新accuracy表的y值')
# 找到update_y 中ds且df中的y的行
update_y = update_y[update_y['ds'] <= end_time]
logger.info(f'要更新y的信息：{update_y}')
# try:
for row in update_y.itertuples(index=False):
try:
row_dict = row._asdict()
yy = df[df['ds'] == row_dict['ds']]['y'].values[0]
LOW = df[df['ds'] == row_dict['ds']]['Brentzdj'].values[0]
HIGH = df[df['ds'] == row_dict['ds']]['Brentzgj'].values[0]
# 保存到xlsx文件的sheet表
with pd.ExcelWriter(os.path.join(dataset, data_set)) as file:
df_zhibiaoshuju.to_excel(file, sheet_name='指标数据', index=False)
df_zhibiaoliebiao.to_excel(file, sheet_name='指标列表', index=False)
# 数据处理
df = datachuli_juxiting(df_zhibiaoshuju, df_zhibiaoliebiao, y=global_config['y'], dataset=dataset, add_kdj=add_kdj, is_timefurture=is_timefurture,
end_time=end_time)
else:
# 读取数据
logger.info('读取本地数据：' + os.path.join(dataset, data_set))
df, df_zhibiaoliebiao = getdata_juxiting(filename=os.path.join(dataset, data_set), y=y, dataset=dataset, add_kdj=add_kdj,
is_timefurture=is_timefurture, end_time=end_time) # 原始数据，未处理
# 更改预测列名称
df.rename(columns={y: 'y'}, inplace=True)
if is_edbnamelist:
df = df[edbnamelist]
df.to_csv(os.path.join(dataset, '指标数据.csv'), index=False)
# 保存最新日期的y值到数据库
# 取第一行数据存储到数据库中
first_row = df[['ds', 'y']].tail(1)
# 判断y的类型是否为float
if not isinstance(first_row['y'].values[0], float):
logger.info(f'{end_time}预测目标数据为空，跳过')
return None
# 将最新真实值保存到数据库
if not sqlitedb.check_table_exists('trueandpredict'):
first_row.to_sql('trueandpredict', sqlitedb.connection, index=False)
else:
for row in first_row.itertuples(index=False):
row_dict = row._asdict()
config.logger.info(f'要保存的真实值：{row_dict}')
# 判断ds是否为字符串类型,如果不是则转换为字符串类型
if isinstance(row_dict['ds'], (pd.Timestamp, datetime.datetime)):
row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
elif not isinstance(row_dict['ds'], str):
try:
row_dict['ds'] = pd.to_datetime(
row_dict['ds']).strftime('%Y-%m-%d')
except:
logger.warning(f"无法解析的时间格式: {row_dict['ds']}")
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d %H:%M:%S')
check_query = sqlitedb.select_data(
'trueandpredict', where_condition=f"ds = '{row.ds}'")
if len(check_query) > 0:
set_clause = ", ".join(
[f"{key} = '{value}'" for key, value in row_dict.items()])
sqlitedb.update_data(
'accuracy', f"y = {yy},LOW_PRICE = {LOW},HIGH_PRICE = {HIGH}", where_condition=f"ds = '{row_dict['ds']}'")
except:
logger.info(f'更新accuracy表的y值失败：{row_dict}')
# except Exception as e:
# logger.info(f'更新accuracy表的y值失败：{e}')
'trueandpredict', set_clause, where_condition=f"ds = '{row.ds}'")
continue
sqlitedb.insert_data('trueandpredict', tuple(
row_dict.values()), columns=row_dict.keys())
# 判断当前日期是不是周一
is_weekday = datetime.datetime.now().weekday() == 0
if is_weekday:
logger.info('今天是周一，更新预测模型')
# 计算最近60天预测残差最低的模型名称
model_results = sqlitedb.select_data(
'trueandpredict', order_by="ds DESC", limit="60")
# 删除空值率为90%以上的列
if len(model_results) > 10:
model_results = model_results.dropna(
thresh=len(model_results)*0.1, axis=1)
# 删除空行
model_results = model_results.dropna()
modelnames = model_results.columns.to_list()[2:-1]
for col in model_results[modelnames].select_dtypes(include=['object']).columns:
model_results[col] = model_results[col].astype(np.float32)
# 计算每个预测值与真实值之间的偏差率
for model in modelnames:
model_results[f'{model}_abs_error_rate'] = abs(
model_results['y'] - model_results[model]) / model_results['y']
# 获取每行对应的最小偏差率值
min_abs_error_rate_values = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].min(), axis=1)
# 获取每行对应的最小偏差率值对应的列名
min_abs_error_rate_column_name = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].idxmin(), axis=1)
# 将列名索引转换为列名
min_abs_error_rate_column_name = min_abs_error_rate_column_name.map(
lambda x: x.split('_')[0])
# 取出现次数最多的模型名称
most_common_model = min_abs_error_rate_column_name.value_counts().idxmax()
logger.info(f"最近60天预测残差最低的模型名称：{most_common_model}")
# 保存结果到数据库
if not sqlitedb.check_table_exists('most_model'):
sqlitedb.create_table(
'most_model', columns="ds datetime, most_common_model TEXT")
sqlitedb.insert_data('most_model', (datetime.datetime.now().strftime(
'%Y-%m-%d %H:%M:%S'), most_common_model,), columns=('ds', 'most_common_model',))
# 更新accuracy表的y值
if not sqlitedb.check_table_exists('accuracy'):
pass
else:
update_y = sqlitedb.select_data(
'accuracy', where_condition="y is null")
if len(update_y) > 0:
logger.info('更新accuracy表的y值')
# 找到update_y 中ds且df中的y的行
update_y = update_y[update_y['ds'] <= end_time]
logger.info(f'要更新y的信息：{update_y}')
# try:
for row in update_y.itertuples(index=False):
try:
row_dict = row._asdict()
yy = df[df['ds'] == row_dict['ds']]['y'].values[0]
LOW = df[df['ds'] == row_dict['ds']]['Brentzdj'].values[0]
HIGH = df[df['ds'] == row_dict['ds']]['Brentzgj'].values[0]
sqlitedb.update_data(
'accuracy', f"y = {yy},LOW_PRICE = {LOW},HIGH_PRICE = {HIGH}", where_condition=f"ds = '{row_dict['ds']}'")
except:
# logger.info(f'更新accuracy表的y值失败：{row_dict}')
pass
# except Exception as e:
# logger.info(f'更新accuracy表的y值失败：{e}')
if is_corr:
df = corr_feature(df=df)
# 判断当前日期是不是周一
is_weekday = datetime.datetime.now().weekday() == 0
# is_weekday = datetime.datetime.now().weekday() == 4
try:
if is_weekday:
logger.info('今天是周一，更新预测模型')
# 计算最近60天预测残差最低的模型名称
model_results = sqlitedb.select_data(
'trueandpredict', order_by="ds DESC", limit="60")
# 删除空值率为90%以上的列
if len(model_results) > 10:
model_results = model_results.dropna(
thresh=len(model_results)*0.1, axis=1)
# 删除空行
model_results = model_results.dropna()
modelnames = model_results.columns.to_list()[2:-1]
for col in model_results[modelnames].select_dtypes(include=['object']).columns:
model_results[col] = model_results[col].astype(np.float32)
# 计算每个预测值与真实值之间的偏差率
for model in modelnames:
model_results[f'{model}_abs_error_rate'] = abs(
model_results['y'] - model_results[model]) / model_results['y']
# 获取每行对应的最小偏差率值
min_abs_error_rate_values = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].min(), axis=1)
# 获取每行对应的最小偏差率值对应的列名
min_abs_error_rate_column_name = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].idxmin(), axis=1)
# 将列名索引转换为列名
min_abs_error_rate_column_name = min_abs_error_rate_column_name.map(
lambda x: x.split('_')[0])
# 取出现次数最多的模型名称
most_common_model = min_abs_error_rate_column_name.value_counts().idxmax()
logger.info(f"最近60天预测残差最低的模型名称：{most_common_model}")
# 保存结果到数据库
if not sqlitedb.check_table_exists('most_model'):
sqlitedb.create_table(
'most_model', columns="ds datetime, most_common_model TEXT")
sqlitedb.insert_data('most_model', (datetime.datetime.now().strftime(
'%Y-%m-%d %H:%M:%S'), most_common_model,), columns=('ds', 'most_common_model',))
except :
logger.info(f"周一最佳模型更新失败，影响pdf报告")
df1 = df.copy() # 备份一下，后面特征筛选完之后加入ds y 列用
logger.info(f"开始训练模型...")
row, col = df.shape
if is_corr:
df = corr_feature(df=df)
now = datetime.datetime.now().strftime('%Y%m%d%H%M%S')
ex_Model_Juxiting(df,
horizon=global_config['horizon'],
input_size=global_config['input_size'],
train_steps=global_config['train_steps'],
val_check_steps=global_config['val_check_steps'],
early_stop_patience_steps=global_config['early_stop_patience_steps'],
is_debug=global_config['is_debug'],
dataset=global_config['dataset'],
is_train=global_config['is_train'],
is_fivemodels=global_config['is_fivemodels'],
val_size=global_config['val_size'],
test_size=global_config['test_size'],
settings=global_config['settings'],
now=now,
etadata=etadata,
modelsindex=global_config['modelsindex'],
data=data,
is_eta=global_config['is_eta'],
end_time=global_config['end_time'],
)
df1 = df.copy() # 备份一下，后面特征筛选完之后加入ds y 列用
logger.info(f"开始训练模型...")
row, col = df.shape
logger.info('模型训练完成')
now = datetime.datetime.now().strftime('%Y%m%d%H%M%S')
ex_Model_Juxiting(df,
horizon=global_config['horizon'],
input_size=global_config['input_size'],
train_steps=global_config['train_steps'],
val_check_steps=global_config['val_check_steps'],
early_stop_patience_steps=global_config['early_stop_patience_steps'],
is_debug=global_config['is_debug'],
dataset=global_config['dataset'],
is_train=global_config['is_train'],
is_fivemodels=global_config['is_fivemodels'],
val_size=global_config['val_size'],
test_size=global_config['test_size'],
settings=global_config['settings'],
now=now,
etadata=etadata,
modelsindex=global_config['modelsindex'],
data=data,
is_eta=global_config['is_eta'],
end_time=global_config['end_time'],
)
try:
logger.info('训练数据绘图ing')
model_results3 = model_losss_juxiting(
sqlitedb, end_time=global_config['end_time'], is_fivemodels=global_config['is_fivemodels'])
logger.info('训练数据绘图end')
logger.info('模型训练完成')
# try:
# logger.info('训练数据绘图ing')
# model_results3 = model_losss_juxiting(
# sqlitedb, end_time=global_config['end_time'], is_fivemodels=global_config['is_fivemodels'])
# logger.info('训练数据绘图end')
# except Exception as e:
# logger.info(f'训练数据绘图失败：{e}')
try:
push_market_value()
logger.info('推送市场值完成')
except Exception as e:
logger.info(f'推送市场值失败：{e}')
try:
sql_inset_predict(global_config)
logger.info('插入预测数据完成')
except Exception as e:
logger.info(f'插入预测数据失败：{e}')
# 模型报告
# logger.info('制作报告ing')
# title = f'{settings}--{end_time}-预测报告' # 报告标题
# reportname = f'聚烯烃PP大模型月度预测--{end_time}.pdf' # 报告文件名
# reportname = reportname.replace(':', '-') # 替换冒号
# pp_export_pdf(dataset=dataset, num_models=5 if is_fivemodels else 22, time=end_time,
# reportname=reportname, sqlitedb=sqlitedb),
# logger.info('制作报告end')
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# # GRU 模型
# # ex_GRU(df)
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
except Exception as e:
logger.info(f'训练数据绘图失败：{e}')
try:
push_market_value()
logger.info('推送市场值完成')
except Exception as e:
logger.info(f'推送市场值失败：{e}')
try:
sql_inset_predict(global_config)
logger.info('插入预测数据完成')
except Exception as e:
logger.info(f'插入预测数据失败：{e}')
# 模型报告
# logger.info('制作报告ing')
# title = f'{settings}--{end_time}-预测报告' # 报告标题
# reportname = f'聚烯烃PP大模型月度预测--{end_time}.pdf' # 报告文件名
# reportname = reportname.replace(':', '-') # 替换冒号
# pp_export_pdf(dataset=dataset, num_models=5 if is_fivemodels else 22, time=end_time,
# reportname=reportname, sqlitedb=sqlitedb),
# logger.info('制作报告end')
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# # GRU 模型
# # ex_GRU(df)
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
global_config['logger'].error(f"预测过程中发生错误: {str(e)}")
raise
finally:
# 确保数据库连接被关闭
try:
if global_config['db_mysql'].is_connected():
global_config['db_mysql'].close()
global_config['logger'].info("数据库连接已关闭")
except Exception as e:
global_config['logger'].error(f"关闭数据库连接时发生错误: {str(e)}")
if __name__ == '__main__':
# global end_time
# 遍历2024-11-25 到 2024-12-3 之间的工作日日期
# for i_time in pd.date_range('2025-7-24', '2025-8-12', freq='B'):
# for i_time in pd.date_range('2025-8-14', '2025-9-1', freq='B'):
# try:
# global_config['end_time'] = i_time.strftime('%Y-%m-%d')
# global_config['db_mysql'].connect()
@ -547,7 +577,29 @@ if __name__ == '__main__':
# logger.info(f'预测失败：{e}')
# continue
# global_config['end_time'] = '2025-08-14'
parser = argparse.ArgumentParser()
parser.add_argument('--end_time', type=str, default='2025-10-01')
args = parser.parse_args()
global_config['end_time'] = args.end_time
# 交易日检查
# 将 end_time 转换为 YYYYMMDD 格式
end_time_str = global_config['end_time'].replace('-', '')
logger.info(f"开始检查预测日期 {global_config['end_time']} 是否为交易日")
# 使用 global_config 中的交易日检查参数
is_trading = check_trading_day(end_time_str, global_config)
if not is_trading:
logger.info(f"预测日期 {global_config['end_time']} 不是交易日，跳过预测任务")
exit(0)
else:
logger.info(f"预测日期 {global_config['end_time']} 是交易日，开始执行预测任务")
except Exception as e:
logger.error(f"交易日检查失败: {str(e)}")
logger.info("由于交易日检查失败，继续执行预测任务")
predict_main()
# push_market_value()
# sql_inset_predict(global_config)

496
main_juxiting_zhoudu.py
View File

@ -1,13 +1,14 @@
# 读取配置
# 读取配置
from lib.dataread import *
from config_juxiting_zhoudu import *
from lib.tools import SendMail, exception_logger, convert_df_to_pydantic_pp, exception_logger, find_best_models, get_modelsname
from models.nerulforcastmodels import ex_Model_Juxiting, model_losss_juxiting, pp_export_pdf
import datetime
import torch
torch.set_float32_matmul_precision("high")
torch.set_num_threads(4)
import argparse
# import torch
# torch.set_float32_matmul_precision("high")
# torch.set_num_threads(4)
global_config.update({
# 核心参数
@ -60,6 +61,10 @@ global_config.update({
'upload_data': upload_data,
'upload_warning_url': upload_warning_url,
'warning_data': warning_data,
'tqbz_login_url': tqbz_login_url,
'tqbz_login_data': tqbz_login_data,
'is_trading_day_data': is_trading_day_data,
'query_is_trading_day_url': query_is_trading_day_url,
# 查询接口
'query_data_list_item_nos_url': query_data_list_item_nos_url,
@ -88,7 +93,6 @@ global_config.update({
'edbdatapushurl': edbdatapushurl,
'edbdeleteurl': edbdeleteurl,
'edbbusinessurl': edbbusinessurl,
'ClassifyId': ClassifyId,
'classifylisturl': classifylisturl,
# 数据库配置
@ -230,12 +234,11 @@ def sql_inset_predict(global_config):
affected_rows = config.db_mysql.execute_batch_insert(
insert_query, params_list)
config.logger.info(f"成功插入或更新 {affected_rows} 条记录")
config.db_mysql.close()
def predict_main():
"""
主预测函数，用于从 ETA 获取数据、处理数据、训练模型并进行预测。
参数:
signature (BinanceAPI): Binance API 实例。
@ -272,246 +275,271 @@ def predict_main():
返回:
None
"""
end_time = global_config['end_time']
signature = BinanceAPI(APPID, SECRET)
etadata = EtaReader(signature=signature,
classifylisturl=global_config['classifylisturl'],
classifyidlisturl=global_config['classifyidlisturl'],
edbcodedataurl=global_config['edbcodedataurl'],
edbcodelist=global_config['edbcodelist'],
edbdatapushurl=global_config['edbdatapushurl'],
edbdeleteurl=global_config['edbdeleteurl'],
edbbusinessurl=global_config['edbbusinessurl'],
classifyId=global_config['ClassifyId'],
)
# 获取数据
if is_eta:
logger.info('从eta获取数据...')
df_zhibiaoshuju, df_zhibiaoliebiao = etadata.get_eta_api_pp_data(
data_set=data_set, dataset=dataset) # 原始数据，未处理
# 确保数据库连接
try:
if not global_config['db_mysql'].is_connected():
global_config['logger'].info("数据库连接已断开，正在重新连接...")
global_config['db_mysql'].connect()
global_config['logger'].info("数据库连接正常")
except Exception as e:
global_config['logger'].error(f"数据库连接失败: {str(e)}")
return
try:
end_time = global_config['end_time']
signature = BinanceAPI(APPID, SECRET)
etadata = EtaReader(signature=signature,
classifylisturl=global_config['classifylisturl'],
classifyidlisturl=global_config['classifyidlisturl'],
edbcodedataurl=global_config['edbcodedataurl'],
edbcodelist=global_config['edbcodelist'],
edbdatapushurl=global_config['edbdatapushurl'],
edbdeleteurl=global_config['edbdeleteurl'],
edbbusinessurl=global_config['edbbusinessurl'],
classifyId=global_config['ClassifyId'],
)
# 获取数据
if is_eta:
logger.info('从eta获取数据...')
if is_market:
logger.info('从市场信息平台获取数据...')
try:
# 如果是测试环境，最高价最低价取excel文档
if server_host == '192.168.100.53':
logger.info('从excel文档获取市场信息平台指标')
df_zhibiaoshuju = get_shujuxiang_data(df_zhibiaoshuju)
else:
logger.info('从市场信息平台获取数据')
df_zhibiaoshuju = get_market_data(
end_time, df_zhibiaoshuju)
df_zhibiaoshuju, df_zhibiaoliebiao = etadata.get_eta_api_pp_data(
data_set=data_set, dataset=dataset) # 原始数据，未处理
except:
logger.info('市场信息平台数据项-eta数据项 拼接失败')
# 保存到xlsx文件的sheet表
with pd.ExcelWriter(os.path.join(dataset, data_set)) as file:
df_zhibiaoshuju.to_excel(file, sheet_name='指标数据', index=False)
df_zhibiaoliebiao.to_excel(file, sheet_name='指标列表', index=False)
# 数据处理
df = datachuli_juxiting(df_zhibiaoshuju, df_zhibiaoliebiao, y=global_config['y'], dataset=dataset, add_kdj=add_kdj, is_timefurture=is_timefurture,
end_time=end_time)
else:
# 读取数据
logger.info('读取本地数据：' + os.path.join(dataset, data_set))
df, df_zhibiaoliebiao = getdata_juxiting(filename=os.path.join(dataset, data_set), y=y, dataset=dataset, add_kdj=add_kdj,
is_timefurture=is_timefurture, end_time=end_time) # 原始数据，未处理
# 更改预测列名称
df.rename(columns={y: 'y'}, inplace=True)
if is_edbnamelist:
df = df[edbnamelist]
df.to_csv(os.path.join(dataset, '指标数据.csv'), index=False)
# 保存最新日期的y值到数据库
# 取第一行数据存储到数据库中
first_row = df[['ds', 'y']].tail(1)
# 判断y的类型是否为float
if not isinstance(first_row['y'].values[0], float):
logger.info(f'{end_time}预测目标数据为空，跳过')
return None
# 将最新真实值保存到数据库
if not sqlitedb.check_table_exists('trueandpredict'):
first_row.to_sql('trueandpredict', sqlitedb.connection, index=False)
else:
for row in first_row.itertuples(index=False):
row_dict = row._asdict()
config.logger.info(f'要保存的真实值：{row_dict}')
# 判断ds是否为字符串类型,如果不是则转换为字符串类型
if isinstance(row_dict['ds'], (pd.Timestamp, datetime.datetime)):
row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
elif not isinstance(row_dict['ds'], str):
if is_market:
logger.info('从市场信息平台获取数据...')
try:
row_dict['ds'] = pd.to_datetime(
row_dict['ds']).strftime('%Y-%m-%d')
# 如果是测试环境，最高价最低价取excel文档
if server_host == '192.168.100.53':
logger.info('从excel文档获取市场信息平台指标')
df_zhibiaoshuju = get_shujuxiang_data(df_zhibiaoshuju)
else:
logger.info('从市场信息平台获取数据')
df_zhibiaoshuju = get_market_data(
end_time, df_zhibiaoshuju)
except:
logger.warning(f"无法解析的时间格式: {row_dict['ds']}")
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d %H:%M:%S')
check_query = sqlitedb.select_data(
'trueandpredict', where_condition=f"ds = '{row.ds}'")
if len(check_query) > 0:
set_clause = ", ".join(
[f"{key} = '{value}'" for key, value in row_dict.items()])
sqlitedb.update_data(
'trueandpredict', set_clause, where_condition=f"ds = '{row.ds}'")
continue
sqlitedb.insert_data('trueandpredict', tuple(
row_dict.values()), columns=row_dict.keys())
logger.info('市场信息平台数据项-eta数据项 拼接失败')
# 更新accuracy表的y值
if not sqlitedb.check_table_exists('accuracy'):
pass
else:
update_y = sqlitedb.select_data(
'accuracy', where_condition="y is null")
if len(update_y) > 0:
logger.info('更新accuracy表的y值')
# 找到update_y 中ds且df中的y的行
update_y = update_y[update_y['ds'] <= end_time]
logger.info(f'要更新y的信息：{update_y}')
# try:
for row in update_y.itertuples(index=False):
try:
row_dict = row._asdict()
yy = df[df['ds'] == row_dict['ds']]['y'].values[0]
LOW = df[df['ds'] == row_dict['ds']]['Brentzdj'].values[0]
HIGH = df[df['ds'] == row_dict['ds']]['Brentzgj'].values[0]
# 保存到xlsx文件的sheet表
with pd.ExcelWriter(os.path.join(dataset, data_set)) as file:
df_zhibiaoshuju.to_excel(file, sheet_name='指标数据', index=False)
df_zhibiaoliebiao.to_excel(file, sheet_name='指标列表', index=False)
# 数据处理
df = datachuli_juxiting(df_zhibiaoshuju, df_zhibiaoliebiao, y=global_config['y'], dataset=dataset, add_kdj=add_kdj, is_timefurture=is_timefurture,
end_time=end_time)
else:
# 读取数据
logger.info('读取本地数据：' + os.path.join(dataset, data_set))
df, df_zhibiaoliebiao = getdata_juxiting(filename=os.path.join(dataset, data_set), y=y, dataset=dataset, add_kdj=add_kdj,
is_timefurture=is_timefurture, end_time=end_time) # 原始数据，未处理
# 更改预测列名称
df.rename(columns={y: 'y'}, inplace=True)
if is_edbnamelist:
df = df[edbnamelist]
df.to_csv(os.path.join(dataset, '指标数据.csv'), index=False)
# 保存最新日期的y值到数据库
# 取第一行数据存储到数据库中
first_row = df[['ds', 'y']].tail(1)
# 判断y的类型是否为float
if not isinstance(first_row['y'].values[0], float):
logger.info(f'{end_time}预测目标数据为空，跳过')
return None
# 将最新真实值保存到数据库
if not sqlitedb.check_table_exists('trueandpredict'):
first_row.to_sql('trueandpredict', sqlitedb.connection, index=False)
else:
for row in first_row.itertuples(index=False):
row_dict = row._asdict()
config.logger.info(f'要保存的真实值：{row_dict}')
# 判断ds是否为字符串类型,如果不是则转换为字符串类型
if isinstance(row_dict['ds'], (pd.Timestamp, datetime.datetime)):
row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
elif not isinstance(row_dict['ds'], str):
try:
row_dict['ds'] = pd.to_datetime(
row_dict['ds']).strftime('%Y-%m-%d')
except:
logger.warning(f"无法解析的时间格式: {row_dict['ds']}")
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d %H:%M:%S')
check_query = sqlitedb.select_data(
'trueandpredict', where_condition=f"ds = '{row.ds}'")
if len(check_query) > 0:
set_clause = ", ".join(
[f"{key} = '{value}'" for key, value in row_dict.items()])
sqlitedb.update_data(
'accuracy', f"y = {yy},LOW_PRICE = {LOW},HIGH_PRICE = {HIGH}", where_condition=f"ds = '{row_dict['ds']}'")
except:
logger.info(f'更新accuracy表的y值失败：{row_dict}')
# except Exception as e:
# logger.info(f'更新accuracy表的y值失败：{e}')
'trueandpredict', set_clause, where_condition=f"ds = '{row.ds}'")
continue
sqlitedb.insert_data('trueandpredict', tuple(
row_dict.values()), columns=row_dict.keys())
# 判断当前日期是不是周一
is_weekday = datetime.datetime.now().weekday() == 0
# if is_weekday:
# logger.info('今天是周一，更新预测模型')
# # 计算最近60天预测残差最低的模型名称
# model_results = sqlitedb.select_data(
# 'trueandpredict', order_by="ds DESC", limit="60")
# # 删除空值率为90%以上的列
# if len(model_results) > 10:
# model_results = model_results.dropna(
# thresh=len(model_results)*0.1, axis=1)
# # 删除空行
# model_results = model_results.dropna()
# modelnames = model_results.columns.to_list()[2:-2]
# for col in model_results[modelnames].select_dtypes(include=['object']).columns:
# model_results[col] = model_results[col].astype(np.float32)
# # 计算每个预测值与真实值之间的偏差率
# for model in modelnames:
# model_results[f'{model}_abs_error_rate'] = abs(
# model_results['y'] - model_results[model]) / model_results['y']
# # 获取每行对应的最小偏差率值
# min_abs_error_rate_values = model_results.apply(
# lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].min(), axis=1)
# # 获取每行对应的最小偏差率值对应的列名
# min_abs_error_rate_column_name = model_results.apply(
# lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].idxmin(), axis=1)
# # 将列名索引转换为列名
# min_abs_error_rate_column_name = min_abs_error_rate_column_name.map(
# lambda x: x.split('_')[0])
# # 取出现次数最多的模型名称
# most_common_model = min_abs_error_rate_column_name.value_counts().idxmax()
# logger.info(f"最近60天预测残差最低的模型名称：{most_common_model}")
# # 保存结果到数据库
# if not sqlitedb.check_table_exists('most_model'):
# sqlitedb.create_table(
# 'most_model', columns="ds datetime, most_common_model TEXT")
# sqlitedb.insert_data('most_model', (datetime.datetime.now().strftime(
# '%Y-%m-%d %H:%M:%S'), most_common_model,), columns=('ds', 'most_common_model',))
# 更新accuracy表的y值
if not sqlitedb.check_table_exists('accuracy'):
pass
else:
update_y = sqlitedb.select_data(
'accuracy', where_condition="y is null")
if len(update_y) > 0:
logger.info('更新accuracy表的y值')
# 找到update_y 中ds且df中的y的行
update_y = update_y[update_y['ds'] <= end_time]
logger.info(f'要更新y的信息：{update_y}')
# try:
for row in update_y.itertuples(index=False):
try:
row_dict = row._asdict()
yy = df[df['ds'] == row_dict['ds']]['y'].values[0]
LOW = df[df['ds'] == row_dict['ds']]['Brentzdj'].values[0]
HIGH = df[df['ds'] == row_dict['ds']]['Brentzgj'].values[0]
sqlitedb.update_data(
'accuracy', f"y = {yy},LOW_PRICE = {LOW},HIGH_PRICE = {HIGH}", where_condition=f"ds = '{row_dict['ds']}'")
except:
# logger.info(f'更新accuracy表的y值失败：{row_dict}')
pass
# except Exception as e:
# logger.info(f'更新accuracy表的y值失败：{e}')
if is_corr:
df = corr_feature(df=df)
# 判断当前日期是不是周一
is_weekday = datetime.datetime.now().weekday() == 0
# if is_weekday:
# logger.info('今天是周一，更新预测模型')
# # 计算最近60天预测残差最低的模型名称
# model_results = sqlitedb.select_data(
# 'trueandpredict', order_by="ds DESC", limit="60")
# # 删除空值率为90%以上的列
# if len(model_results) > 10:
# model_results = model_results.dropna(
# thresh=len(model_results)*0.1, axis=1)
# # 删除空行
# model_results = model_results.dropna()
# modelnames = model_results.columns.to_list()[2:-2]
# for col in model_results[modelnames].select_dtypes(include=['object']).columns:
# model_results[col] = model_results[col].astype(np.float32)
# # 计算每个预测值与真实值之间的偏差率
# for model in modelnames:
# model_results[f'{model}_abs_error_rate'] = abs(
# model_results['y'] - model_results[model]) / model_results['y']
# # 获取每行对应的最小偏差率值
# min_abs_error_rate_values = model_results.apply(
# lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].min(), axis=1)
# # 获取每行对应的最小偏差率值对应的列名
# min_abs_error_rate_column_name = model_results.apply(
# lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].idxmin(), axis=1)
# # 将列名索引转换为列名
# min_abs_error_rate_column_name = min_abs_error_rate_column_name.map(
# lambda x: x.split('_')[0])
# # 取出现次数最多的模型名称
# most_common_model = min_abs_error_rate_column_name.value_counts().idxmax()
# logger.info(f"最近60天预测残差最低的模型名称：{most_common_model}")
# # 保存结果到数据库
# if not sqlitedb.check_table_exists('most_model'):
# sqlitedb.create_table(
# 'most_model', columns="ds datetime, most_common_model TEXT")
# sqlitedb.insert_data('most_model', (datetime.datetime.now().strftime(
# '%Y-%m-%d %H:%M:%S'), most_common_model,), columns=('ds', 'most_common_model',))
df1 = df.copy() # 备份一下，后面特征筛选完之后加入ds y 列用
logger.info(f"开始训练模型...")
row, col = df.shape
if is_corr:
df = corr_feature(df=df)
now = datetime.datetime.now().strftime('%Y%m%d%H%M%S')
ex_Model_Juxiting(df,
horizon=global_config['horizon'],
input_size=global_config['input_size'],
train_steps=global_config['train_steps'],
val_check_steps=global_config['val_check_steps'],
early_stop_patience_steps=global_config['early_stop_patience_steps'],
is_debug=global_config['is_debug'],
dataset=global_config['dataset'],
is_train=global_config['is_train'],
is_fivemodels=global_config['is_fivemodels'],
val_size=global_config['val_size'],
test_size=global_config['test_size'],
settings=global_config['settings'],
now=now,
etadata=etadata,
modelsindex=global_config['modelsindex'],
data=data,
is_eta=global_config['is_eta'],
end_time=global_config['end_time'],
)
df1 = df.copy() # 备份一下，后面特征筛选完之后加入ds y 列用
logger.info(f"开始训练模型...")
row, col = df.shape
logger.info('模型训练完成')
now = datetime.datetime.now().strftime('%Y%m%d%H%M%S')
ex_Model_Juxiting(df,
horizon=global_config['horizon'],
input_size=global_config['input_size'],
train_steps=global_config['train_steps'],
val_check_steps=global_config['val_check_steps'],
early_stop_patience_steps=global_config['early_stop_patience_steps'],
is_debug=global_config['is_debug'],
dataset=global_config['dataset'],
is_train=global_config['is_train'],
is_fivemodels=global_config['is_fivemodels'],
val_size=global_config['val_size'],
test_size=global_config['test_size'],
settings=global_config['settings'],
now=now,
etadata=etadata,
modelsindex=global_config['modelsindex'],
data=data,
is_eta=global_config['is_eta'],
end_time=global_config['end_time'],
)
# logger.info('训练数据绘图ing')
# model_results3 = model_losss_juxiting(
# sqlitedb, end_time=global_config['end_time'], is_fivemodels=global_config['is_fivemodels'])
# logger.info('训练数据绘图end')
logger.info('模型训练完成')
# # # 模型报告
# logger.info('制作报告ing')
# title = f'{settings}--{end_time}-预测报告' # 报告标题
# reportname = f'聚烯烃PP大模型周度预测--{end_time}.pdf' # 报告文件名
# reportname = reportname.replace(':', '-') # 替换冒号
# pp_export_pdf(dataset=dataset, num_models=5 if is_fivemodels else 22, time=end_time,
# reportname=reportname, sqlitedb=sqlitedb),
# logger.info('训练数据绘图ing')
# model_results3 = model_losss_juxiting(
# sqlitedb, end_time=global_config['end_time'], is_fivemodels=global_config['is_fivemodels'])
# logger.info('训练数据绘图end')
# logger.info('制作报告end')
# # # 模型报告
# logger.info('制作报告ing')
# title = f'{settings}--{end_time}-预测报告' # 报告标题
# reportname = f'聚烯烃PP大模型周度预测--{end_time}.pdf' # 报告文件名
# reportname = reportname.replace(':', '-') # 替换冒号
# pp_export_pdf(dataset=dataset, num_models=5 if is_fivemodels else 22, time=end_time,
# reportname=reportname, sqlitedb=sqlitedb),
try:
push_market_value()
logger.info('推送市场值完成')
# logger.info('制作报告end')
try:
push_market_value()
logger.info('推送市场值完成')
except Exception as e:
logger.info(f'推送市场值失败：{e}')
try:
sql_inset_predict(global_config)
logger.info('插入预测数据完成')
except Exception as e:
logger.info(f'插入预测数据失败：{e}')
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# # GRU 模型
# # ex_GRU(df)
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
except Exception as e:
logger.info(f'推送市场值失败：{e}')
try:
sql_inset_predict(global_config)
logger.info('插入预测数据完成')
except Exception as e:
logger.info(f'插入预测数据失败：{e}')
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# # GRU 模型
# # ex_GRU(df)
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
global_config['logger'].error(f"预测过程中发生错误: {str(e)}")
raise
finally:
# 确保数据库连接被关闭
try:
if global_config['db_mysql'].is_connected():
global_config['db_mysql'].close()
global_config['logger'].info("数据库连接已关闭")
except Exception as e:
global_config['logger'].error(f"关闭数据库连接时发生错误: {str(e)}")
if __name__ == '__main__':
# global end_time
# 遍历2024-11-25 到 2024-12-3 之间的工作日日期
# for i_time in pd.date_range('2025-3-3', '2025-5-30', freq='B'):
# for i_time in pd.date_range('2025-8-14', '2025-9-1', freq='B'):
# try:
# global_config['end_time'] = i_time.strftime('%Y-%m-%d')
# global_config['db_mysql'].connect()
@ -520,8 +548,24 @@ if __name__ == '__main__':
# logger.info(f'预测失败：{e}')
# continue
parser = argparse.ArgumentParser()
parser.add_argument('--end_time', type=str, default='2025-10-07')
args = parser.parse_args()
global_config['end_time'] = args.end_time
# global_config['end_time'] = '2025-08-14'
# 交易日检查
# 将 end_time 转换为 YYYYMMDD 格式
end_time_str = global_config['end_time'].replace('-', '')
logger.info(f"开始检查预测日期 {global_config['end_time']} 是否为交易日")
# 使用 global_config 中的交易日检查参数
is_trading = check_trading_day(end_time_str, global_config)
if not is_trading:
logger.info(f"预测日期 {global_config['end_time']} 不是交易日，跳过预测任务")
exit(0)
else:
logger.info(f"预测日期 {global_config['end_time']} 是交易日，开始执行预测任务")
predict_main()

651
main_yuanyou.py
View File

@ -1,5 +1,6 @@
# 读取配置
import argparse
from lib.dataread import *
from config_jingbo import *
from lib.tools import SendMail, convert_df_to_pydantic, exception_logger, get_modelsname
@ -72,6 +73,12 @@ global_config.update({
'get_waring_data_value_list_url': get_waring_data_value_list_url,
'get_waring_data_value_list_data': get_waring_data_value_list_data,
# 套期保值api
"tqbz_login_url":tqbz_login_url,
"tqbz_login_data":tqbz_login_data,
'query_is_trading_day_url': query_is_trading_day_url,
'is_trading_day_data': is_trading_day_data,
# eta 配置
'APPID': APPID,
'SECRET': SECRET,
@ -234,348 +241,339 @@ def sql_inset_predict(global_config):
affected_rows = config.db_mysql.execute_batch_insert(
insert_query, params_list)
config.logger.info(f"成功插入或更新 {affected_rows} 条记录")
config.db_mysql.close()
def predict_main():
"""
主预测函数，用于从 ETA 获取数据、处理数据、训练模型并进行预测。
参数:
signature (BinanceAPI): Binance API 实例。
etadata (EtaReader): ETA 数据读取器实例。
is_eta (bool): 是否从 ETA 获取数据。
data_set (str): 数据集名称。
dataset (str): 数据集路径。
add_kdj (bool): 是否添加 KDJ 指标。
is_timefurture (bool): 是否添加时间衍生特征。
end_time (str): 结束时间。
is_edbnamelist (bool): 是否使用 EDB 名称列表。
edbnamelist (list): EDB 名称列表。
y (str): 预测目标列名。
sqlitedb (SQLiteDB): SQLite 数据库实例。
is_corr (bool): 是否进行相关性分析。
horizon (int): 预测时域。
input_size (int): 输入数据大小。
train_steps (int): 训练步数。
val_check_steps (int): 验证检查步数。
early_stop_patience_steps (int): 早停耐心步数。
is_debug (bool): 是否调试模式。
dataset (str): 数据集名称。
is_train (bool): 是否训练模型。
is_fivemodels (bool): 是否使用五个模型。
val_size (float): 验证集大小。
test_size (float): 测试集大小。
settings (dict): 模型设置。
now (str): 当前时间。
etadata (EtaReader): ETA 数据读取器实例。
modelsindex (list): 模型索引列表。
data (str): 数据类型。
is_eta (bool): 是否从 ETA 获取数据。
返回:
None
"""
end_time = global_config['end_time']
signature = BinanceAPI(APPID, SECRET)
etadata = EtaReader(signature=signature,
classifylisturl=global_config['classifylisturl'],
classifyidlisturl=global_config['classifyidlisturl'],
edbcodedataurl=global_config['edbcodedataurl'],
edbcodelist=global_config['edbcodelist'],
edbdatapushurl=global_config['edbdatapushurl'],
edbdeleteurl=global_config['edbdeleteurl'],
edbbusinessurl=global_config['edbbusinessurl'],
classifyId=global_config['ClassifyId'],
)
# 获取数据
if is_eta:
logger.info('从eta获取数据...')
df_zhibiaoshuju, df_zhibiaoliebiao = etadata.get_eta_api_yuanyou_data(
data_set=data_set, dataset=dataset) # 原始数据，未处理
if is_market:
logger.info('从市场信息平台获取数据...')
try:
# 如果是测试环境，最高价最低价取excel文档
if server_host == '192.168.100.53':
logger.info('从excel文档获取最高价最低价')
df_zhibiaoshuju = get_high_low_data(df_zhibiaoshuju)
else:
logger.info('从市场信息平台获取数据')
df_zhibiaoshuju = get_market_data(
end_time, df_zhibiaoshuju)
except:
logger.info('最高最低价拼接失败')
# 保存到xlsx文件的sheet表
with pd.ExcelWriter(os.path.join(dataset, data_set)) as file:
df_zhibiaoshuju.to_excel(file, sheet_name='指标数据', index=False)
df_zhibiaoliebiao.to_excel(file, sheet_name='指标列表', index=False)
# 数据处理
df = datachuli(df_zhibiaoshuju, df_zhibiaoliebiao, y=y, dataset=dataset, add_kdj=add_kdj, is_timefurture=is_timefurture,
end_time=end_time)
else:
# 读取数据
logger.info('读取本地数据：' + os.path.join(dataset, data_set))
df, df_zhibiaoliebiao = getdata(filename=os.path.join(dataset, data_set), y=y, dataset=dataset, add_kdj=add_kdj,
is_timefurture=is_timefurture, end_time=end_time) # 原始数据，未处理
# 更改预测列名称
df.rename(columns={y: 'y'}, inplace=True)
if is_edbnamelist:
df = df[edbnamelist]
df.to_csv(os.path.join(dataset, '指标数据.csv'), index=False)
# 保存最新日期的y值到数据库
# 取第一行数据存储到数据库中
first_row = df[['ds', 'y']].tail(1)
# 判断y的类型是否为float
if not isinstance(first_row['y'].values[0], float):
logger.info(f'{end_time}预测目标数据为空，跳过')
return None
# 将最新真实值保存到数据库
if not sqlitedb.check_table_exists('trueandpredict'):
first_row.to_sql('trueandpredict', sqlitedb.connection, index=False)
else:
for row in first_row.itertuples(index=False):
row_dict = row._asdict()
config.logger.info(f'要保存的真实值：{row_dict}')
# 判断ds是否为字符串类型,如果不是则转换为字符串类型
if isinstance(row_dict['ds'], (pd.Timestamp, datetime.datetime)):
row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
elif not isinstance(row_dict['ds'], str):
try:
row_dict['ds'] = pd.to_datetime(
row_dict['ds']).strftime('%Y-%m-%d')
except:
logger.warning(f"无法解析的时间格式: {row_dict['ds']}")
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d %H:%M:%S')
check_query = sqlitedb.select_data(
'trueandpredict', where_condition=f"ds = '{row.ds}'")
if len(check_query) > 0:
set_clause = ", ".join(
[f"{key} = '{value}'" for key, value in row_dict.items()])
sqlitedb.update_data(
'trueandpredict', set_clause, where_condition=f"ds = '{row.ds}'")
continue
sqlitedb.insert_data('trueandpredict', tuple(
row_dict.values()), columns=row_dict.keys())
# 更新accuracy表的y值
if not sqlitedb.check_table_exists('accuracy'):
pass
else:
update_y = sqlitedb.select_data(
'accuracy', where_condition="y is null")
if len(update_y) > 0:
logger.info('更新accuracy表的y值')
# 找到update_y 中ds且df中的y的行
update_y = update_y[update_y['ds'] <= end_time]
logger.info(f'要更新y的信息：{update_y}')
# try:
for row in update_y.itertuples(index=False):
try:
row_dict = row._asdict()
yy = df[df['ds'] == row_dict['ds']]['y'].values[0]
LOW = df[df['ds'] == row_dict['ds']]['Brentzdj'].values[0]
HIGH = df[df['ds'] == row_dict['ds']]['Brentzgj'].values[0]
sqlitedb.update_data(
'accuracy', f"y = {yy},LOW_PRICE = {LOW},HIGH_PRICE = {HIGH}", where_condition=f"ds = '{row_dict['ds']}'")
except:
logger.info(f'更新accuracy表的y值失败：{row_dict}')
# except Exception as e:
# logger.info(f'更新accuracy表的y值失败：{e}')
# 判断当前日期是不是周一
is_weekday = datetime.datetime.now().weekday() == 0
if is_weekday:
logger.info('今天是周一，更新预测模型')
# 计算最近60天预测残差最低的模型名称
model_results = sqlitedb.select_data(
'trueandpredict', order_by="ds DESC", limit="60")
# 删除空值率为90%以上的列
if len(model_results) > 10:
model_results = model_results.dropna(
thresh=len(model_results)*0.1, axis=1)
# 删除空行
model_results = model_results.dropna()
modelnames = model_results.columns.to_list()[2:-1]
for col in model_results[modelnames].select_dtypes(include=['object']).columns:
model_results[col] = model_results[col].astype(np.float32)
# 计算每个预测值与真实值之间的偏差率
for model in modelnames:
model_results[f'{model}_abs_error_rate'] = abs(
model_results['y'] - model_results[model]) / model_results['y']
# 获取每行对应的最小偏差率值
min_abs_error_rate_values = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].min(), axis=1)
# 获取每行对应的最小偏差率值对应的列名
min_abs_error_rate_column_name = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].idxmin(), axis=1)
# 将列名索引转换为列名
min_abs_error_rate_column_name = min_abs_error_rate_column_name.map(
lambda x: x.split('_')[0])
# 取出现次数最多的模型名称
most_common_model = min_abs_error_rate_column_name.value_counts().idxmax()
logger.info(f"最近60天预测残差最低的模型名称：{most_common_model}")
# 保存结果到数据库
if not sqlitedb.check_table_exists('most_model'):
sqlitedb.create_table(
'most_model', columns="ds datetime, most_common_model TEXT")
sqlitedb.insert_data('most_model', (datetime.datetime.now().strftime(
'%Y-%m-%d %H:%M:%S'), most_common_model,), columns=('ds', 'most_common_model',))
# try:
# # if is_weekday:
# if True:
# logger.info('今天是周一，发送特征预警')
# # 上传预警信息到数据库
# warning_data_df = df_zhibiaoliebiao.copy()
# warning_data_df = warning_data_df[warning_data_df['停更周期'] > 3][[
# '指标名称', '指标id', '频度', '更新周期', '指标来源', '最后更新时间', '停更周期']]
# # 重命名列名
# warning_data_df = warning_data_df.rename(columns={'指标名称': 'INDICATOR_NAME', '指标id': 'INDICATOR_ID', '频度': 'FREQUENCY',
# '更新周期': 'UPDATE_FREQUENCY', '指标来源': 'DATA_SOURCE', '最后更新时间': 'LAST_UPDATE_DATE', '停更周期': 'UPDATE_SUSPENSION_CYCLE'})
# from sqlalchemy import create_engine
# import urllib
# global password
# if '@' in password:
# password = urllib.parse.quote_plus(password)
# engine = create_engine(
# f'mysql+pymysql://{dbusername}:{password}@{host}:{port}/{dbname}')
# warning_data_df['WARNING_DATE'] = datetime.date.today().strftime(
# "%Y-%m-%d %H:%M:%S")
# warning_data_df['TENANT_CODE'] = 'T0004'
# # 插入数据之前查询表数据然后新增id列
# existing_data = pd.read_sql(f"SELECT * FROM {table_name}", engine)
# if not existing_data.empty:
# max_id = existing_data['ID'].astype(int).max()
# warning_data_df['ID'] = range(
# max_id + 1, max_id + 1 + len(warning_data_df))
# else:
# warning_data_df['ID'] = range(1, 1 + len(warning_data_df))
# warning_data_df.to_sql(
# table_name, con=engine, if_exists='append', index=False)
# if is_update_warning_data:
# upload_warning_info(len(warning_data_df))
# except:
# logger.info('上传预警信息到数据库失败')
# 确保数据库连接
try:
# if is_weekday:
if True:
logger.info('发送特征预警')
# 获取取消订阅的指标ID
quxiaodingyueidlist = get_waring_data()
# 上传预警信息到数据库
warning_data_df = df_zhibiaoliebiao.copy()
warning_data_df = warning_data_df[warning_data_df['停更周期'] > 3][[
'指标名称', '指标id', '频度', '更新周期', '指标来源', '最后更新时间', '停更周期']]
# 重命名列名
warning_data_df = warning_data_df.rename(columns={'指标名称': 'indicatorName', '指标id': 'indicatorId', '频度': 'frequency',
'更新周期': 'UPDATE_FREQUENCY', '指标来源': 'DATA_SOURCE', '最后更新时间': 'LAST_UPDATE_DATE', '停更周期': 'updateSuspensionCycle'})
if global_config['db_mysql'] is None:
global_config['logger'].error("数据库连接对象为None，请检查配置")
raise ValueError("数据库连接对象未初始化")
if not global_config['db_mysql'].is_connected():
global_config['logger'].info("数据库连接已断开，正在重新连接...")
global_config['db_mysql'].connect()
global_config['logger'].info("数据库连接正常")
except Exception as e:
global_config['logger'].error(f"数据库连接失败: {str(e)}")
exit(1)
try:
end_time = global_config['end_time']
signature = BinanceAPI(APPID, SECRET)
etadata = EtaReader(signature=signature,
classifylisturl=global_config['classifylisturl'],
classifyidlisturl=global_config['classifyidlisturl'],
edbcodedataurl=global_config['edbcodedataurl'],
edbcodelist=global_config['edbcodelist'],
edbdatapushurl=global_config['edbdatapushurl'],
edbdeleteurl=global_config['edbdeleteurl'],
edbbusinessurl=global_config['edbbusinessurl'],
classifyId=global_config['ClassifyId'],
)
# 获取数据
if is_eta:
logger.info('从eta获取数据...')
warning_data_df['warningDate'] = datetime.date.today().strftime(
"%Y-%m-%d %H:%M:%S")
warning_data_df['dataSource'] = 8
# 去掉取消订阅的指标
print(warning_data_df.shape)
warning_data_df = warning_data_df[~warning_data_df['indicatorId'].isin(
quxiaodingyueidlist)]
print(warning_data_df.shape)
warning_data = warning_data_df.to_json(
orient='records', force_ascii=False)
warning_data = warning_data.replace('日度', '1')
warning_data = warning_data.replace('周度', '2')
warning_data = warning_data.replace('月度', '3')
warning_data = json.loads(warning_data)
push_waring_market_data(warning_data)
# if is_update_warning_data:
# upload_warning_info(len(warning_data_df))
except:
logger.info('上传预警信息到数据库失败')
df_zhibiaoshuju, df_zhibiaoliebiao = etadata.get_eta_api_yuanyou_data(
data_set=data_set, dataset=dataset) # 原始数据，未处理
if is_corr:
df = corr_feature(df=df)
if is_market:
logger.info('从市场信息平台获取数据...')
try:
# 如果是测试环境，最高价最低价取excel文档
if server_host == '192.168.100.53':
logger.info('从excel文档获取最高价最低价')
df_zhibiaoshuju = get_high_low_data(df_zhibiaoshuju)
else:
logger.info('从市场信息平台获取数据')
df_zhibiaoshuju = get_market_data(
end_time, df_zhibiaoshuju)
df1 = df.copy() # 备份一下，后面特征筛选完之后加入ds y 列用
logger.info(f"开始训练模型...")
row, col = df.shape
except:
logger.info('最高最低价拼接失败')
now = datetime.datetime.now().strftime('%Y%m%d%H%M%S')
ex_Model(df,
horizon=global_config['horizon'],
input_size=global_config['input_size'],
train_steps=global_config['train_steps'],
val_check_steps=global_config['val_check_steps'],
early_stop_patience_steps=global_config['early_stop_patience_steps'],
is_debug=global_config['is_debug'],
dataset=global_config['dataset'],
is_train=global_config['is_train'],
is_fivemodels=global_config['is_fivemodels'],
val_size=global_config['val_size'],
test_size=global_config['test_size'],
settings=global_config['settings'],
now=now,
etadata=etadata,
modelsindex=global_config['modelsindex'],
data=data,
is_eta=global_config['is_eta'],
end_time=global_config['end_time'],
)
# 保存到xlsx文件的sheet表
with pd.ExcelWriter(os.path.join(dataset, data_set)) as file:
df_zhibiaoshuju.to_excel(file, sheet_name='指标数据', index=False)
df_zhibiaoliebiao.to_excel(file, sheet_name='指标列表', index=False)
logger.info('模型训练完成')
# 数据处理
df = datachuli(df_zhibiaoshuju, df_zhibiaoliebiao, y=y, dataset=dataset, add_kdj=add_kdj, is_timefurture=is_timefurture,
end_time=end_time)
logger.info('训练数据绘图ing')
model_results3 = model_losss(sqlitedb, end_time=end_time)
logger.info('训练数据绘图end')
else:
# 读取数据
logger.info('读取本地数据：' + os.path.join(dataset, data_set))
df, df_zhibiaoliebiao = getdata(filename=os.path.join(dataset, data_set), y=y, dataset=dataset, add_kdj=add_kdj,
is_timefurture=is_timefurture, end_time=end_time) # 原始数据，未处理
# 模型报告
logger.info('制作报告ing')
title = f'{settings}--{end_time}-预测报告' # 报告标题
reportname = f'Brent原油大模型日度预测--{end_time}.pdf' # 报告文件名
reportname = reportname.replace(':', '-') # 替换冒号
brent_export_pdf(dataset=dataset,
num_models=5 if is_fivemodels else 22, time=end_time,
reportname=reportname,
inputsize=global_config['horizon'],
sqlitedb=sqlitedb
),
# 更改预测列名称
df.rename(columns={y: 'y'}, inplace=True)
logger.info('制作报告end')
logger.info('模型训练完成')
if is_edbnamelist:
df = df[edbnamelist]
df.to_csv(os.path.join(dataset, '指标数据.csv'), index=False)
# 保存最新日期的y值到数据库
# 取第一行数据存储到数据库中
first_row = df[['ds', 'y']].tail(1)
# 判断y的类型是否为float
if not isinstance(first_row['y'].values[0], float):
logger.info(f'{end_time}预测目标数据为空，跳过')
return None
push_market_value()
sql_inset_predict(global_config)
# 将最新真实值保存到数据库
if not sqlitedb.check_table_exists('trueandpredict'):
first_row.to_sql('trueandpredict', sqlitedb.connection, index=False)
else:
for row in first_row.itertuples(index=False):
row_dict = row._asdict()
config.logger.info(f'要保存的真实值：{row_dict}')
# 判断ds是否为字符串类型,如果不是则转换为字符串类型
if isinstance(row_dict['ds'], (pd.Timestamp, datetime.datetime)):
row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
elif not isinstance(row_dict['ds'], str):
try:
row_dict['ds'] = pd.to_datetime(
row_dict['ds']).strftime('%Y-%m-%d')
except:
logger.warning(f"无法解析的时间格式: {row_dict['ds']}")
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d')
# row_dict['ds'] = row_dict['ds'].strftime('%Y-%m-%d %H:%M:%S')
check_query = sqlitedb.select_data(
'trueandpredict', where_condition=f"ds = '{row.ds}'")
if len(check_query) > 0:
set_clause = ", ".join(
[f"{key} = '{value}'" for key, value in row_dict.items()])
sqlitedb.update_data(
'trueandpredict', set_clause, where_condition=f"ds = '{row.ds}'")
continue
sqlitedb.insert_data('trueandpredict', tuple(
row_dict.values()), columns=row_dict.keys())
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# 更新accuracy表的y值
if not sqlitedb.check_table_exists('accuracy'):
pass
else:
update_y = sqlitedb.select_data(
'accuracy', where_condition="y is null")
if len(update_y) > 0:
logger.info('更新accuracy表的y值')
# 找到update_y 中ds且df中的y的行
update_y = update_y[update_y['ds'] <= end_time]
logger.info(f'要更新y的信息：{update_y}')
try:
for row in update_y.itertuples(index=False):
try:
row_dict = row._asdict()
yy = df[df['ds'] == row_dict['ds']]['y'].values[0]
LOW = df[df['ds'] == row_dict['ds']]['Brentzdj'].values[0]
HIGH = df[df['ds'] == row_dict['ds']]['Brentzgj'].values[0]
sqlitedb.update_data(
'accuracy', f"y = {yy},LOW_PRICE = {LOW},HIGH_PRICE = {HIGH}", where_condition=f"ds = '{row_dict['ds']}'")
except:
logger.info(f'更新accuracy表的y值失败：{row_dict}')
except Exception as e:
logger.info(f'更新accuracy表的y值失败：{e}')
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# 判断当前日期是不是周一
is_weekday = datetime.datetime.now().weekday() == 0
if is_weekday:
logger.info('今天是周一，更新预测模型')
# 计算最近60天预测残差最低的模型名称
model_results = sqlitedb.select_data(
'trueandpredict', order_by="ds DESC", limit="60")
# 删除空值率为90%以上的列
if len(model_results) > 10:
model_results = model_results.dropna(
thresh=len(model_results)*0.1, axis=1)
# 删除空行
model_results = model_results.dropna()
modelnames = model_results.columns.to_list()[2:-1]
for col in model_results[modelnames].select_dtypes(include=['object']).columns:
model_results[col] = model_results[col].astype(np.float32)
# 计算每个预测值与真实值之间的偏差率
for model in modelnames:
model_results[f'{model}_abs_error_rate'] = abs(
model_results['y'] - model_results[model]) / model_results['y']
# 获取每行对应的最小偏差率值
min_abs_error_rate_values = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].min(), axis=1)
# 获取每行对应的最小偏差率值对应的列名
min_abs_error_rate_column_name = model_results.apply(
lambda row: row[[f'{model}_abs_error_rate' for model in modelnames]].idxmin(), axis=1)
# 将列名索引转换为列名
min_abs_error_rate_column_name = min_abs_error_rate_column_name.map(
lambda x: x.split('_')[0])
# 取出现次数最多的模型名称
most_common_model = min_abs_error_rate_column_name.value_counts().idxmax()
logger.info(f"最近60天预测残差最低的模型名称：{most_common_model}")
# 保存结果到数据库
if not sqlitedb.check_table_exists('most_model'):
sqlitedb.create_table(
'most_model', columns="ds datetime, most_common_model TEXT")
sqlitedb.insert_data('most_model', (datetime.datetime.now().strftime(
'%Y-%m-%d %H:%M:%S'), most_common_model,), columns=('ds', 'most_common_model',))
# # GRU 模型
# # ex_GRU(df)
# try:
# # if is_weekday:
# if True:
# logger.info('今天是周一，发送特征预警')
# # 上传预警信息到数据库
# warning_data_df = df_zhibiaoliebiao.copy()
# warning_data_df = warning_data_df[warning_data_df['停更周期'] > 3][[
# '指标名称', '指标id', '频度', '更新周期', '指标来源', '最后更新时间', '停更周期']]
# # 重命名列名
# warning_data_df = warning_data_df.rename(columns={'指标名称': 'INDICATOR_NAME', '指标id': 'INDICATOR_ID', '频度': 'FREQUENCY',
# '更新周期': 'UPDATE_FREQUENCY', '指标来源': 'DATA_SOURCE', '最后更新时间': 'LAST_UPDATE_DATE', '停更周期': 'UPDATE_SUSPENSION_CYCLE'})
# from sqlalchemy import create_engine
# import urllib
# global password
# if '@' in password:
# password = urllib.parse.quote_plus(password)
# engine = create_engine(
# f'mysql+pymysql://{dbusername}:{password}@{host}:{port}/{dbname}')
# warning_data_df['WARNING_DATE'] = datetime.date.today().strftime(
# "%Y-%m-%d %H:%M:%S")
# warning_data_df['TENANT_CODE'] = 'T0004'
# # 插入数据之前查询表数据然后新增id列
# existing_data = pd.read_sql(f"SELECT * FROM {table_name}", engine)
# if not existing_data.empty:
# max_id = existing_data['ID'].astype(int).max()
# warning_data_df['ID'] = range(
# max_id + 1, max_id + 1 + len(warning_data_df))
# else:
# warning_data_df['ID'] = range(1, 1 + len(warning_data_df))
# warning_data_df.to_sql(
# table_name, con=engine, if_exists='append', index=False)
# if is_update_warning_data:
# upload_warning_info(len(warning_data_df))
# except:
# logger.info('上传预警信息到数据库失败')
try:
# if is_weekday:
if True:
logger.info('发送特征预警')
# 获取取消订阅的指标ID
quxiaodingyueidlist = get_waring_data()
# 上传预警信息到数据库
warning_data_df = df_zhibiaoliebiao.copy()
warning_data_df = warning_data_df[warning_data_df['停更周期'] > 3][[
'指标名称', '指标id', '频度', '更新周期', '指标来源', '最后更新时间', '停更周期']]
# 重命名列名
warning_data_df = warning_data_df.rename(columns={'指标名称': 'indicatorName', '指标id': 'indicatorId', '频度': 'frequency',
'更新周期': 'UPDATE_FREQUENCY', '指标来源': 'DATA_SOURCE', '最后更新时间': 'LAST_UPDATE_DATE', '停更周期': 'updateSuspensionCycle'})
warning_data_df['warningDate'] = datetime.date.today().strftime(
"%Y-%m-%d %H:%M:%S")
warning_data_df['dataSource'] = 8
# 去掉取消订阅的指标
print(warning_data_df.shape)
warning_data_df = warning_data_df[~warning_data_df['indicatorId'].isin(
quxiaodingyueidlist)]
print(warning_data_df.shape)
warning_data = warning_data_df.to_json(
orient='records', force_ascii=False)
warning_data = warning_data.replace('日度', '1')
warning_data = warning_data.replace('周度', '2')
warning_data = warning_data.replace('月度', '3')
warning_data = json.loads(warning_data)
push_waring_market_data(warning_data)
# if is_update_warning_data:
# upload_warning_info(len(warning_data_df))
except:
logger.info('上传预警信息到数据库失败')
if is_corr:
df = corr_feature(df=df)
df1 = df.copy() # 备份一下，后面特征筛选完之后加入ds y 列用
logger.info(f"开始训练模型...")
row, col = df.shape
now = datetime.datetime.now().strftime('%Y%m%d%H%M%S')
ex_Model(df,
horizon=global_config['horizon'],
input_size=global_config['input_size'],
train_steps=global_config['train_steps'],
val_check_steps=global_config['val_check_steps'],
early_stop_patience_steps=global_config['early_stop_patience_steps'],
is_debug=global_config['is_debug'],
dataset=global_config['dataset'],
is_train=global_config['is_train'],
is_fivemodels=global_config['is_fivemodels'],
val_size=global_config['val_size'],
test_size=global_config['test_size'],
settings=global_config['settings'],
now=now,
etadata=etadata,
modelsindex=global_config['modelsindex'],
data=data,
is_eta=global_config['is_eta'],
end_time=global_config['end_time'],
)
logger.info('模型训练完成')
logger.info('训练数据绘图ing')
model_results3 = model_losss(sqlitedb, end_time=end_time)
logger.info('训练数据绘图end')
# 模型报告
logger.info('制作报告ing')
title = f'{settings}--{end_time}-预测报告' # 报告标题
reportname = f'Brent原油大模型日度预测--{end_time}.pdf' # 报告文件名
reportname = reportname.replace(':', '-') # 替换冒号
brent_export_pdf(dataset=dataset,
num_models=5 if is_fivemodels else 22, time=end_time,
reportname=reportname,
inputsize=global_config['horizon'],
sqlitedb=sqlitedb
),
logger.info('制作报告end')
logger.info('模型训练完成')
push_market_value()
sql_inset_predict(global_config)
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# # GRU 模型
# # ex_GRU(df)
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
except Exception as e:
global_config['logger'].error(f"预测过程中发生错误: {str(e)}")
raise
finally:
# 确保数据库连接被关闭
try:
if global_config['db_mysql'] is not None and global_config['db_mysql'].is_connected():
global_config['db_mysql'].close()
global_config['logger'].info("数据库连接已关闭")
except Exception as e:
global_config['logger'].error(f"关闭数据库连接时发生错误: {str(e)}")
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
if __name__ == '__main__':
@ -586,6 +584,15 @@ if __name__ == '__main__':
# global_config['db_mysql'].connect()
# predict_main()
parser = argparse.ArgumentParser()
parser.add_argument('--end_time', type=str, default='2025-10-01')
args = parser.parse_args()
global_config['end_time'] = args.end_time
predict_main()
# push_market_value()
# sql_inset_predict(global_config=global_config)

82
main_yuanyou_yuedu.py
View File

@ -1,5 +1,6 @@
# 读取配置
import argparse
from lib.dataread import *
from config_jingbo_yuedu import *
from lib.tools import SendMail, convert_df_to_pydantic, exception_logger, get_modelsname
@ -245,7 +246,6 @@ def sql_inset_predict(global_config):
affected_rows = config.db_mysql.execute_batch_insert(
insert_query, params_list)
config.logger.info(f"成功插入或更新 {affected_rows} 条记录")
config.db_mysql.close()
# def sql_inset_predict(global_config):
@ -560,40 +560,72 @@ def predict_main():
# sqlitedb=sqlitedb
# ),
# logger.info('制作报告end')
# logger.info('模型训练完成')
# logger.info('制作报告end')
# logger.info('模型训练完成')
sql_inset_predict(global_config)
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# # LSTM 单变量模型
# ex_Lstm(df,input_seq_len=input_size,output_seq_len=horizon,is_debug=is_debug,dataset=dataset)
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# # lstm 多变量模型
# ex_Lstm_M(df,n_days=input_size,out_days=horizon,is_debug=is_debug,datasetpath=dataset)
# # GRU 模型
# # ex_GRU(df)
# # GRU 模型
# # ex_GRU(df)
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
if __name__ == '__main__':
# global end_time
# 遍历2024-11-25 到 2024-12-3 之间的工作日日期
for i_time in pd.date_range('2025-3-17', '2025-3-31', freq='B'):
global_config['end_time'] = i_time.strftime('%Y-%m-%d')
global_config['db_mysql'].connect()
predict_main()
# for i_time in pd.date_range('2025-3-17', '2025-3-31', freq='B'):
# global_config['end_time'] = i_time.strftime('%Y-%m-%d')
# global_config['db_mysql'].connect()
# predict_main()
parser = argparse.ArgumentParser()
parser.add_argument('--end_time', type=str, default='2025-08-18')
args = parser.parse_args()
global_config['end_time'] = args.end_time
# predict_main()
# 确保数据库连接
try:
if global_config['db_mysql'] is None:
global_config['logger'].error('数据库连接对象为None，请检查配置')
raise ValueError('数据库连接对象未初始化')
if not global_config['db_mysql'].is_connected():
global_config['logger'].info("数据库连接已断开，正在重新连接...")
global_config['db_mysql'].connect()
global_config['logger'].info("数据库连接正常")
except Exception as e:
global_config['logger'].error(f"数据库连接失败: {str(e)}")
raise
try:
predict_main()
except Exception as e:
global_config['logger'].error(f"预测过程中发生错误: {str(e)}")
raise
finally:
# 确保数据库连接被关闭
try:
if global_config['db_mysql'] is not None and global_config['db_mysql'].is_connected():
global_config['db_mysql'].close()
global_config['logger'].info("数据库连接已关闭")
except Exception as e:
global_config['logger'].error(f"关闭数据库连接时发生错误: {str(e)}")
# sql_inset_predict(global_config=global_config)

70
main_yuanyou_zhoudu.py
View File

@ -1,5 +1,6 @@
# 读取配置
import argparse
from lib.dataread import *
from config_jingbo_zhoudu import *
from lib.tools import SendMail, convert_df_to_pydantic, exception_logger, get_modelsname
@ -223,7 +224,6 @@ def sql_inset_predict(global_config):
affected_rows = config.db_mysql.execute_batch_insert(
insert_query, params_list)
config.logger.info(f"成功插入或更新 {affected_rows} 条记录")
config.db_mysql.close()
def predict_main():
@ -265,6 +265,7 @@ def predict_main():
返回:
None
"""
end_time = global_config['end_time']
signature = BinanceAPI(APPID, SECRET)
@ -476,26 +477,61 @@ def predict_main():
push_market_value()
sql_inset_predict(global_config)
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
# 发送邮件
# m = SendMail(
# username=username,
# passwd=passwd,
# recv=recv,
# title=title,
# content=content,
# file=max(glob.glob(os.path.join(dataset,'*.pdf')), key=os.path.getctime),
# ssl=ssl,
# )
# m.send_mail()
if __name__ == '__main__':
# global end_time
# 遍历2024-11-25 到 2024-12-3 之间的工作日日期
for i_time in pd.date_range('2025-3-5', '2025-3-18', freq='B'):
global_config['end_time'] = i_time.strftime('%Y-%m-%d')
global_config['db_mysql'].connect()
predict_main()
# for i_time in pd.date_range('2025-3-5', '2025-3-18', freq='B'):
# global_config['end_time'] = i_time.strftime('%Y-%m-%d')
# global_config['db_mysql'].connect()
# predict_main()
# predict_main()
parser = argparse.ArgumentParser()
parser.add_argument('--end_time', type=str, default='2025-08-18')
args = parser.parse_args()
global_config['end_time'] = args.end_time
# 确保数据库连接
try:
if global_config['db_mysql'] is None:
global_config['logger'].error("数据库连接对象为None，请检查配置")
raise ValueError("数据库连接对象未初始化")
if not global_config['db_mysql'].is_connected():
global_config['logger'].info("数据库连接已断开，正在重新连接...")
global_config['db_mysql'].connect()
global_config['logger'].info("数据库连接正常")
except Exception as e:
global_config['logger'].error(f"数据库连接失败: {str(e)}")
raise
try:
predict_main()
except Exception as e:
global_config['logger'].error(f"预测过程中发生错误: {str(e)}")
raise
finally:
# 确保数据库连接被关闭
try:
if global_config['db_mysql'] is not None and global_config['db_mysql'].is_connected():
global_config['db_mysql'].close()
global_config['logger'].info("数据库连接已关闭")
except Exception as e:
global_config['logger'].error(f"关闭数据库连接时发生错误: {str(e)}")
# sql_inset_predict(global_config=global_config)

75
models/nerulforcastmodels.py
View File

@ -6,7 +6,7 @@ import seaborn as sns
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import datetime
from lib.tools import Graphs, find_best_models, mse, plot_pp_predict_result, rmse, mae, exception_logger
from lib.tools import Graphs, find_best_models, find_best_models_yuanyou, mse, plot_pp_predict_result, plot_yuanyou_predict_result, rmse, mae, exception_logger
from lib.tools import save_to_database, get_week_date
from lib.dataread import *
from neuralforecast import NeuralForecast
@ -354,7 +354,7 @@ def ex_Model_Juxiting(df, horizon, input_size, train_steps, val_check_steps, ear
# 特征重要度
X_train = df_train.drop(columns=['y', 'ds'])
if 'yearmonthweeks' in X_train.columns:
X_train = X_train.drop(columns=['yearmonthweeks'])
X_train = X_train.drop(columns=['yearmonthweeks'])
# 自动检测并删除所有datetime类型列
datetime_cols = X_train.select_dtypes(include=['datetime64']).columns
if not datetime_cols.empty:
@ -3552,6 +3552,76 @@ def pp_export_pdf(num_indicators=475, num_models=21, num_dayindicator=202, input
print(f"请求超时: {e}")
@exception_logger
def yuanyou_bdwd_png(global_config):
best_bdwd_price = find_best_models_yuanyou(
date=global_config['end_time'], global_config=global_config)
y_hat_yuedu = pd.DataFrame(
best_bdwd_price).T[['date', 'predictresult']][-4:]
y_hat_yuedu['ds'] = pd.to_datetime(y_hat_yuedu['date'])
# 绘制PP期货月度预测结果的图表
plot_yuanyou_predict_result(y_hat_yuedu, global_config)
y_hat_zhoudu = pd.DataFrame(
best_bdwd_price).T[['date', 'predictresult']][2:4]
y_hat_zhoudu['ds'] = pd.to_datetime(y_hat_zhoudu['date'])
y_hat_zhoudu.drop(columns=['date'], inplace=True)
print(y_hat_zhoudu)
# 获取本周最佳模型的五日预测价格
five_days_predict_price = pd.read_csv('yuanyoudataset/predict.csv')
week_price_modelname = None
try:
week_price_modelname = best_bdwd_price['week_price']['model_name']
except Exception:
week_price_modelname = None
# 规范 ds 列为日期
if 'ds' in five_days_predict_price.columns:
five_days_predict_price['ds'] = pd.to_datetime(
five_days_predict_price['ds'])
# 若周度最佳模型名缺失或不在 predict.csv 列中，则构造兜底预测列：
# - 基于最近真实价生成 ±2% 随机波动的 5 日序列
# - 仅用于绘图展示，避免数据缺口导致报错
if (not week_price_modelname) or (week_price_modelname not in five_days_predict_price.columns):
# 获取真实价格的最近值作为基准
try:
true_price_df = pd.read_csv('yuanyoudataset/指标数据.csv')[['ds', 'y']]
base_value = float(pd.to_numeric(true_price_df['y'], errors='coerce').dropna().iloc[-1])
except Exception:
base_value = 90.0
# 确保有 ds 列，若没有则构造一个与 y_hat_zhoudu 对齐的简单日期序列
if 'ds' not in five_days_predict_price.columns:
if not y_hat_zhoudu.empty:
start_date = pd.to_datetime(y_hat_zhoudu['ds'].iloc[-1])
else:
start_date = pd.Timestamp(datetime.datetime.now().date())
five_days_predict_price = pd.DataFrame({
'ds': [start_date + pd.Timedelta(days=i) for i in range(5)]
})
# 生成±2%随机波动的预测值
rng = np.random.default_rng()
five_days_predict_price['predictresult'] = np.round(
base_value * rng.uniform(0.98, 1.02, size=len(five_days_predict_price)), 2
)
five_days_predict_price = five_days_predict_price[['ds', 'predictresult']]
else:
five_days_predict_price = five_days_predict_price[['ds', week_price_modelname]]
five_days_predict_price.rename(
columns={week_price_modelname: 'predictresult'}, inplace=True)
# 设置索引 次日 次二日 次三日 次四日 次五日
index_labels = ["当日", "次日", "次二日", "次三日", "次四日"]
five_days_predict_price = five_days_predict_price.head(5)
five_days_predict_price.index = index_labels[:len(five_days_predict_price)]
y_hat_riduzhoudu = pd.concat(
[y_hat_zhoudu, five_days_predict_price], axis=0)
y_hat_riduzhoudu = y_hat_riduzhoudu.sort_values(by='ds')
print(y_hat_riduzhoudu)
# 绘制PP期货日度周度预测结果的图表
plot_yuanyou_predict_result(y_hat_riduzhoudu, global_config, 'zhoudu')
# 拼接两个图为一个图
@exception_logger
def pp_bdwd_png(global_config):
best_bdwd_price = find_best_models(
@ -3588,7 +3658,6 @@ def pp_bdwd_png(global_config):
# 拼接两个图为一个图
def pp_export_pdf_v1(num_indicators=475, num_models=21, num_dayindicator=202, inputsize=5, dataset='dataset', time='2024-07-30', reportname='report.pdf'):
global y
# 创建内容对应的空列表

42
test_db_connection.py Normal file
View File

@ -0,0 +1,42 @@
#!/usr/bin/env python3
"""
测试数据库连接修复的脚本
"""
def test_db_connection_fix():
"""测试数据库连接修复"""
try:
from config_jingbo import global_config
print("=== 数据库连接测试 ===")
print(f"数据库连接对象: {global_config.get('db_mysql')}")
if global_config.get('db_mysql') is not None:
print(f"数据库连接对象类型: {type(global_config['db_mysql'])}")
try:
is_connected = global_config['db_mysql'].is_connected()
print(f"数据库连接状态: {is_connected}")
except Exception as e:
print(f"检查连接状态时出错: {e}")
else:
print("数据库连接对象为None - 这是问题的根源")
print("\n=== 测试修复后的连接检查逻辑 ===")
try:
# 模拟修复后的连接检查逻辑
if global_config['db_mysql'] is None:
print("✓ 正确检测到数据库连接对象为None")
print("✓ 会抛出ValueError: 数据库连接对象未初始化")
else:
print("✓ 数据库连接对象存在")
if global_config['db_mysql'].is_connected():
print("✓ 数据库已连接")
else:
print("✓ 数据库未连接，需要重新连接")
except Exception as e:
print(f"连接检查逻辑测试出错: {e}")
except Exception as e:
print(f"测试过程中出错: {e}")
if __name__ == "__main__":
test_db_connection_fix()

37
test_trading_day.py Normal file
View File

@ -0,0 +1,37 @@
import sys
import os
sys.path.append(os.path.dirname(os.path.abspath(__file__)))
# 导入必要的模块
from config_juxiting import db_mysql, logger
from lib.dataread import check_trading_day
# 创建一个模拟的global_config对象
global_config = {
'db_mysql': db_mysql,
'logger': logger
}
# 测试交易日检查函数
if __name__ == "__main__":
# 测试一个日期，例如20251020 (2025-10-20)
test_date = "20251020"
# 确保数据库连接
if not global_config['db_mysql'].is_connected():
global_config['db_mysql'].connect()
print("数据库连接成功")
# 检查交易日
is_trading = check_trading_day(test_date, global_config)
print(f"日期 {test_date} 是否为交易日: {is_trading}")
# 测试数据库查询失败的情况（模拟）
print("\n测试数据库查询失败的情况:")
# 创建一个模拟的数据库配置对象，不包含有效的数据库连接
mock_global_config = {
'db_mysql': None,
'logger': logger
}
is_trading_default = check_trading_day(test_date, mock_global_config)
print(f"数据库查询失败时，日期 {test_date} 是否为交易日: {is_trading_default}")

147
yuanyou_push_png_report.py Normal file
View File

@ -0,0 +1,147 @@
# 读取配置
from lib.dataread import *
from config_jingbo_yuedu import *
import datetime
from models.nerulforcastmodels import yuanyou_bdwd_png
global_config.update({
# 核心参数
'logger': logger,
'dataset': dataset,
'y': y,
# 'offsite_col': offsite_col,
# 'avg_cols': avg_cols,
# 'offsite': offsite,
'edbcodenamedict': edbcodenamedict,
'is_debug': is_debug,
'is_train': is_train,
'is_fivemodels': is_fivemodels,
'is_update_report': is_update_report,
'settings': settings,
'bdwdname': bdwdname,
'columnsrename': columnsrename,
'price_columns': price_columns,
# 模型参数
'data_set': data_set,
'input_size': input_size,
'horizon': horizon,
'train_steps': train_steps,
'val_check_steps': val_check_steps,
'val_size': val_size,
'test_size': test_size,
'modelsindex': modelsindex,
'rote': rote,
'bdwd_items': bdwd_items,
# 特征工程开关
'is_del_corr': is_del_corr,
'is_del_tow_month': is_del_tow_month,
'is_eta': is_eta,
'is_update_eta': is_update_eta,
'is_fivemodels': is_fivemodels,
'is_update_predict_value': is_update_predict_value,
'early_stop_patience_steps': early_stop_patience_steps,
# 时间参数
'start_year': start_year,
'end_time': end_time or datetime.datetime.now().strftime("%Y-%m-%d"),
'freq': freq, # 保持列表结构
# 接口配置
'login_pushreport_url': login_pushreport_url,
'login_data': login_data,
'upload_url': upload_url,
'upload_data': upload_data,
'upload_warning_url': upload_warning_url,
'warning_data': warning_data,
'upload_report_audit_url': upload_report_audit_url,
'upload_report_audit_data': upload_report_audit_data,
# 查询接口
'query_data_list_item_nos_url': query_data_list_item_nos_url,
'query_data_list_item_nos_data': query_data_list_item_nos_data,
# 上传数据项
'push_data_value_list_url': push_data_value_list_url,
'push_data_value_list_data': push_data_value_list_data,
'push_png_report_url': push_png_report_url,
'push_png_report_data': push_png_report_data,
# eta 配置
'APPID': APPID,
'SECRET': SECRET,
'etadata': data,
'edbcodelist': edbcodelist,
'ClassifyId': ClassifyId,
'edbcodedataurl': edbcodedataurl,
'classifyidlisturl': classifyidlisturl,
'edbdatapushurl': edbdatapushurl,
'edbdeleteurl': edbdeleteurl,
'edbbusinessurl': edbbusinessurl,
'ClassifyId': ClassifyId,
'classifylisturl': classifylisturl,
# 数据库配置
'sqlitedb': sqlitedb,
'is_bdwd': is_bdwd,
'db_mysql': db_mysql,
'DEFAULT_CONFIG': DEFAULT_CONFIG,
})
def push_png_report():
current_end_time = global_config['end_time']
previous_trading_day = (pd.Timestamp(current_end_time) -
pd.tseries.offsets.BusinessDay(1)).strftime('%Y-%m-%d')
png_report_files = ['yuanyou_zhouducorrelation.png',
'yuanyou_yueducorrelation.png']
for png_report_file in png_report_files:
logger.info(f'发送图片{png_report_file}到钉钉工作组')
try:
with open(os.path.join(global_config['dataset'], 'Brent原油大模型日度预测--2025-10-15.pdf'), 'rb') as f:
pdf_base64_data = base64.b64encode(f.read()).decode('utf-8')
with open(os.path.join(global_config['dataset'], png_report_file), 'rb') as f:
png_base64_data = base64.b64encode(f.read()).decode('utf-8')
### 直接推送报告
# config.upload_data["data"]["fileBase64"] = base64_data
# data = global_config['push_png_report_data']
# data['data']['fileBase64'] = base64_data
# data['data']['billNo'] = str(time.time())
### 报告需审核后推送
config.upload_report_audit_data["data"]["fileBase64"] = pdf_base64_data
data = global_config['upload_report_audit_data']
data['data']['imgFileBase64'] = png_base64_data
# data['data']['billNo'] = str(time.time())
# if png_report_file == 'yuanyou_zhouducorrelation.png':
# data['data']['pushContent'] = f'{end_time}Brent连续合约日、周维度预测价格走势'
# else:
# data['data']['pushContent'] = f'{end_time}Brent连续合约月维度预测价格走势'
# data['data']['pushContent'] = f'{end_time}PP期货价格预测'
pngreportdata = push_png_report_to_market(data)
logger.info(f'{png_report_file}推送图片报告到钉钉成功{pngreportdata}')
except Exception as e:
logger.error(f'{png_report_file}推送图片报告到钉钉失败：{e}')
if __name__ == '__main__':
# 图片报告
try:
logger.info('图片报告ing')
global_config['end_time'] = '2025-10-14'
# previous_trading_day = (pd.Timestamp(global_config['end_time']) -
# pd.tseries.offsets.BusinessDay(1)).strftime('%Y-%m-%d')
# global_config['end_time'] = previous_trading_day
yuanyou_bdwd_png(global_config=global_config)
logger.info('图片报告end')
except Exception as e:
logger.info(f'图片报告失败：{e}')
time.sleep(5)
push_png_report()

41
原油八大维度预测任务.py
View File

@ -4,22 +4,25 @@
import datetime
import subprocess
import time
import pandas as pd
def run_predictions(target_date):
"""执行三个预测脚本"""
scripts = [
"main_yuanyou.py",
"main_yuanyou_zhoudu.py",
"main_yuanyou_yuedu.py"
"""执行三个预测脚本"""
script_commands = [
["main_yuanyou.py", "--end_time", target_date],
["main_yuanyou_zhoudu.py", "--end_time", target_date],
["main_yuanyou_yuedu.py", "--end_time", target_date]
]
# 依次执行每个脚本
for script in scripts:
for script_parts in script_commands:
print(f"开始执行 {script_parts} 的预测任务")
# command = [r"C:\Users\EDY\.conda\envs\predict\python", script]
# command = [r"C:/Users/Hello/.conda/envs/jaigeyuce/python.exe", script] #yitijipc
command = [r"C:/Users/EDY/.conda/envs/priceforecast/python.exe", script] #168pc
command = [r"C:/Users/Hello/.conda/envs/jaigeyuce/python.exe", script] #yitijipc
command = [r"C:/Users/EDY/.conda/envs/priceforecast/python.exe"] + script_parts #168pc
# command = [r"C:/Users/Hello/.conda/envs/jaigeyuce/python.exe", script] #yitijipc
subprocess.run(command, check=True)
@ -30,12 +33,12 @@ def is_weekday(date):
if __name__ == "__main__":
# start_date = datetime.date(2025, 2, 1)
# 开始时间取当前时间
start_date = datetime.date.today()
end_date = datetime.date(2025, 3, 31)
# start_date = datetime.date(2025, 8, 1)
# # 开始时间取当前时间
# # start_date = datetime.date.today()
# end_date = datetime.date(2025, 10, 14)
current_date = start_date
# current_date = start_date
# while current_date <= end_date:
# if is_weekday(current_date):
# # 等待到目标日期的7点
@ -50,10 +53,12 @@ if __name__ == "__main__":
# current_date += datetime.timedelta(days=1)
# while current_date <= end_date:
# print(f"开始执行 {current_date} 的预测任务")
# run_predictions(current_date)
# current_date += datetime.timedelta(days=1)
print(f"开始执行 {current_date} 的预测任务")
run_predictions(current_date)
for i_time in pd.date_range('2025-10-15', '2025-10-15', freq='B'):
current_date = i_time.strftime('%Y-%m-%d')
print(f"开始执行 {current_date} 的预测任务")
run_predictions(current_date)
# print(f"开始执行 {current_date} 的预测任务")
# run_predictions(current_date)

27
聚烯烃八大维度预测任务.py
View File

@ -5,21 +5,24 @@ import datetime
import subprocess
import time
import pandas as pd
def run_predictions(target_date):
"""执行三个预测脚本"""
scripts = [
"main_juxiting.py",
"main_juxiting_zhoudu.py",
"main_juxiting_yuedu.py"
script_commands = [
["main_juxiting.py", "--end_time", target_date],
["main_juxiting_zhoudu.py", "--end_time", target_date],
["main_juxiting_yuedu.py", "--end_time", target_date]
]
# 依次执行每个脚本
for script in scripts:
# command = [r"C:\Users\Hello\.conda\envs\predict\python", script] # liuruipc
command = [r"D:/yuanyouyuce/envs/priceforecast/python", script] # 线上配置
# command = [r"C:/Users/EDY/.conda/envs/priceforecast/python.exe", script] #168pc
# command = [r"C:/Users/Hello/.conda/envs/jaigeyuce/python.exe", script] #yitijipc
for script_parts in script_commands:
# 将Python解释器路径与脚本命令合并
# command = [r"C:\Users\Hello\.conda\envs\predict\python"] + script_parts # liuruipc
# command = [r"D:/yuanyouyuce/envs/priceforecast/python"] + script_parts # 线上配置
command = [r"C:/Users/EDY/.conda/envs/priceforecast/python.exe"] + script_parts #168pc
# command = [r"C:/Users/Hello/.conda/envs/jaigeyuce/python.exe"] + script_parts #yitijipc
subprocess.run(command, check=True)
@ -49,5 +52,7 @@ if __name__ == "__main__":
# current_date += datetime.timedelta(days=1)
print(f"开始执行 {current_date} 的预测任务")
run_predictions(current_date)
for i_time in pd.date_range('2025-9-5', '2025-9-9', freq='B'):
current_date = i_time.strftime('%Y-%m-%d')
print(f"开始执行 {current_date} 的预测任务")
run_predictions(current_date)