revert 3a339c5270

revert 在原始代码的基础上，增加（load、water、carbon数据集、处理)，增加flask接口，将test运行的数据，传入到run.py,返回给test。py窗口

wgz_forecast(lh)/carbon/carbon_inference.py

@@ -1,24 +0,0 @@
-import joblib
-import numpy as np
-
-
-def load_carbon_model(path):
-    gbm = joblib.load(path)
-    return gbm
-def carbon_forecast(inputs: np.ndarray, model):
-    """_summary_
-
-    Args:
-        inputs (np.ndarray): 输入序列
-        model (_type_): _description_
-    """
-    out = model.predict([inputs])
-    return out
-
-
-if __name__ == '__main__':
-    model = load_carbon_model('./models/carbon_pred.joblib')
-    inputs = np.random.randn(24)
-    print(inputs.shape)
-    out = carbon_forecast(inputs, model)
-    print(out)

wgz_forecast(lh)/carbon/carbon_train.py

@@ -1,67 +0,0 @@
-import pandas as pd
-import numpy as np
-import xgboost as xgb
-from sklearn.model_selection import train_test_split
-from sklearn.metrics import mean_squared_error, r2_score
-import joblib
-from logzero import logger
-
-def time_series_to_supervised(data, columns, n_in=24, n_out=1,dropnan=True):
-    """
-    :param data:作为列表或2D NumPy数组的观察序列。需要。
-    :param n_in:作为输入的滞后观察数（X）。值可以在[1..len（数据）]之间可选。默认为1。
-    :param n_out:作为输出的观测数量（y）。值可以在[0..len（数据）]之间。可选的。默认为1。
-    :param dropnan:Boolean是否删除具有NaN值的行。可选的。默认为True。
-    :return:
-    """
-    logger.info(f"正在处理训练数据：size：{data.shape}")
-    n_vars = 1 if type(data) is list else data.shape[1]
-    df = pd.DataFrame(data)
-    origNames = columns
-    cols, names = list(), list()
-    cols.append(df.shift(0))
-    names += [('%s' % origNames[j]) for j in range(n_vars)]
-    n_in = max(1, n_in)
-    for i in range(n_in-1, 0, -1):
-        time = '(t-%d)' % i
-        cols.append(df.shift(i))
-        names += [('%s%s' % (origNames[j], time)) for j in range(n_vars)]
-    n_out = max(n_out, 0)
-    for i in range(1, n_out+1):
-        time = '(t+%d)' % i
-        cols.append(df.shift(-i))
-        names += [('%s%s' % (origNames[j], time)) for j in range(n_vars)]
-    agg = pd.concat(cols, axis=1)
-    agg.columns = names
-    if dropnan:
-        agg.dropna(inplace=True)
-    return agg
-
-def train_model(train_data: pd.DataFrame):
-    """训练模型的函数，需要根据模型类型实际调整
-    Args:
-        data (pd.DataFrame): 训练集
-    """
-    # 特征和输出列名，需要根据业务场景灵活处理
-    fea_cols  = train_data.columns[:-1].tolist()
-    out_cols = train_data.columns[-1:].tolist()
-    logger.info(f"特征列: {fea_cols}, 输出列: {out_cols}")
-    X = train_data[fea_cols]
-    y = train_data[out_cols]
-    train_X,test_X,train_y,test_y = train_test_split(X, y, test_size=0.2, random_state=42)
-    valid_X,test_X,valid_y,test_y = train_test_split(test_X, test_y, test_size=0.5, random_state=42)
-    # 参数
-    other_params = {'learning_rate': 0.1, 'n_estimators': 150, 'max_depth': 10, 'min_child_weight': 1, 'seed': 0, 'subsample': 0.8, 'colsample_bytree': 0.8, 'gamma': 0, 'reg_alpha': 0, 'reg_lambda': 1}
-    print(train_X.shape, train_y.shape)
-    gbm = xgb.XGBRegressor(objective='reg:squarederror', early_stopping_rounds=20,**other_params)
-    gbm.fit(train_X.values, train_y.values, eval_set=[(valid_X.values, valid_y.values)])
-    y_pred = gbm.predict(test_X.values)
-    logger.info(f"Root Mean Squared Error on Test set: {np.sqrt(mean_squared_error(test_y, y_pred))}")
-    logger.info(f"R2 score on Test set: {r2_score(test_y, y_pred)}")
-    joblib.dump(gbm, './models/carbon_pred.joblib')
-    logger.info(f"save_path: ./models/carbon_pred.joblib")
-
-if __name__ == '__main__':
-    data = pd.read_csv('./data/carbon_data_hourly.csv', index_col=0)
-    agg = time_series_to_supervised(data.values, data.columns, 24, 1)
-    train_model(agg)

wgz_forecast(lh)/carbon/data/carbon_data_hourly.csv

@@ -1,567 +0,0 @@
-date,Value
-2022-01-01 00:00:00,72.076
-2022-01-01 01:00:00,64.442
-2022-01-01 02:00:00,64.084
-2022-01-01 03:00:00,60.842
-2022-01-01 04:00:00,61.798
-2022-01-01 05:00:00,66.538
-2022-01-01 06:00:00,72.626
-2022-01-01 07:00:00,75.181
-2022-01-01 08:00:00,68.397
-2022-01-01 09:00:00,67.668
-2022-01-01 10:00:00,67.021
-2022-01-01 11:00:00,71.118
-2022-01-01 12:00:00,811.791
-2022-01-01 13:00:00,70.55
-2022-01-01 14:00:00,62.929
-2022-01-01 15:00:00,64.519
-2022-01-01 16:00:00,60.544
-2022-01-01 17:00:00,64.687
-2022-01-01 18:00:00,64.736
-2022-01-01 19:00:00,73.698
-2022-01-01 20:00:00,72.559
-2022-01-01 21:00:00,63.221
-2022-01-01 22:00:00,65.418
-2022-01-01 23:00:00,65.593
-2022-01-02 00:00:00,71.677
-2022-01-02 01:00:00,800.131
-2022-01-02 02:00:00,72.708
-2022-01-02 03:00:00,65.117
-2022-01-02 04:00:00,66.532
-2022-01-02 05:00:00,61.975
-2022-01-02 06:00:00,62.031
-2022-01-02 07:00:00,67.875
-2022-01-02 08:00:00,74.184
-2022-01-02 09:00:00,77.029
-2022-01-02 10:00:00,66.156
-2022-01-02 11:00:00,66.651
-2022-01-02 12:00:00,67.625
-2022-01-02 13:00:00,75.857
-2022-01-02 14:00:00,823.74
-2022-01-02 15:00:00,81.264
-2022-01-02 16:00:00,71.058
-2022-01-02 17:00:00,73.361
-2022-01-02 18:00:00,68.703
-2022-01-02 19:00:00,68.985
-2022-01-02 20:00:00,73.936
-2022-01-02 21:00:00,80.968
-2022-01-02 22:00:00,81.962
-2022-01-02 23:00:00,72.95
-2022-01-03 00:00:00,74.746
-2022-01-03 01:00:00,78.924
-2022-01-03 02:00:00,84.49
-2022-01-03 03:00:00,911.346
-2022-01-03 04:00:00,87.215
-2022-01-03 05:00:00,75.943
-2022-01-03 06:00:00,75.092
-2022-01-03 07:00:00,68.641
-2022-01-03 08:00:00,74.916
-2022-01-03 09:00:00,80.942
-2022-01-03 10:00:00,90.667
-2022-01-03 11:00:00,89.046
-2022-01-03 12:00:00,80.975
-2022-01-03 13:00:00,77.369
-2022-01-03 14:00:00,78.058
-2022-01-03 15:00:00,83.279
-2022-01-03 16:00:00,962.143
-2022-01-03 17:00:00,85.191
-2022-01-03 18:00:00,71.476
-2022-01-03 19:00:00,67.829
-2022-01-03 20:00:00,69.051
-2022-01-03 21:00:00,74.201
-2022-01-03 22:00:00,81.372
-2022-01-03 23:00:00,88.003
-2022-01-04 00:00:00,91.836
-2022-01-04 01:00:00,85.066
-2022-01-04 02:00:00,79.494
-2022-01-04 03:00:00,79.291
-2022-01-04 04:00:00,87.107
-2022-01-04 05:00:00,959.916
-2022-01-04 06:00:00,93.946
-2022-01-04 07:00:00,83.91
-2022-01-04 08:00:00,83.689
-2022-01-04 09:00:00,78.077
-2022-01-04 10:00:00,83.19
-2022-01-04 11:00:00,88.15
-2022-01-04 12:00:00,96.579
-2022-01-04 13:00:00,97.246
-2022-01-04 14:00:00,84.463
-2022-01-04 15:00:00,86.071
-2022-01-04 16:00:00,86.09
-2022-01-04 17:00:00,94.293
-2022-01-04 18:00:00,1055.704
-2022-01-04 19:00:00,100.568
-2022-01-04 20:00:00,94.862
-2022-01-04 21:00:00,93.21
-2022-01-04 22:00:00,81.245
-2022-01-04 23:00:00,82.785
-2022-01-05 00:00:00,91.484
-2022-01-05 01:00:00,107.126
-2022-01-05 02:00:00,106.245
-2022-01-05 03:00:00,95.662
-2022-01-05 04:00:00,90.03
-2022-01-05 05:00:00,91.239
-2022-01-05 06:00:00,102.138
-2022-01-05 07:00:00,1136.594
-2022-01-05 08:00:00,108.164
-2022-01-05 09:00:00,94.766
-2022-01-05 10:00:00,95.724
-2022-01-05 11:00:00,86.386
-2022-01-05 12:00:00,89.002
-2022-01-05 13:00:00,99.05
-2022-01-05 14:00:00,111.044
-2022-01-05 15:00:00,107.758
-2022-01-05 16:00:00,95.891
-2022-01-05 17:00:00,94.541
-2022-01-05 18:00:00,92.986
-2022-01-05 19:00:00,105.056
-2022-01-05 20:00:00,1180.367
-2022-01-05 21:00:00,113.136
-2022-01-05 22:00:00,97.313
-2022-01-05 23:00:00,95.335
-2022-01-06 00:00:00,86.587
-2022-01-06 01:00:00,90.83
-2022-01-06 02:00:00,94.419
-2022-01-06 03:00:00,109.995
-2022-01-06 04:00:00,109.179
-2022-01-06 05:00:00,96.39
-2022-01-06 06:00:00,92.24
-2022-01-06 07:00:00,95.165
-2022-01-06 08:00:00,101.367
-2022-01-06 09:00:00,1181.956
-2022-01-06 10:00:00,105.965
-2022-01-06 11:00:00,90.912
-2022-01-06 12:00:00,93.568
-2022-01-06 13:00:00,86.575
-2022-01-06 14:00:00,90.75
-2022-01-06 15:00:00,99.98
-2022-01-06 16:00:00,119.945
-2022-01-06 17:00:00,126.652
-2022-01-06 18:00:00,107.675
-2022-01-06 19:00:00,100.677
-2022-01-06 20:00:00,101.663
-2022-01-06 21:00:00,117.417
-2022-01-06 22:00:00,1241.778
-2022-01-06 23:00:00,119.436
-2022-01-07 00:00:00,103.635
-2022-01-07 01:00:00,108.149
-2022-01-07 02:00:00,94.329
-2022-01-07 03:00:00,98.181
-2022-01-07 04:00:00,112.983
-2022-01-07 05:00:00,119.701
-2022-01-07 06:00:00,125.724
-2022-01-07 07:00:00,107.179
-2022-01-07 08:00:00,108.583
-2022-01-07 09:00:00,107.546
-2022-01-07 10:00:00,112.167
-2022-01-07 11:00:00,1317.614
-2022-01-07 12:00:00,125.429
-2022-01-07 13:00:00,109.359
-2022-01-07 14:00:00,107.966
-2022-01-07 15:00:00,100.317
-2022-01-07 16:00:00,107.595
-2022-01-07 17:00:00,113.584
-2022-01-07 18:00:00,126.626
-2022-01-07 19:00:00,124.309
-2022-01-07 20:00:00,111.9
-2022-01-07 21:00:00,108.331
-2022-01-07 22:00:00,107.034
-2022-01-07 23:00:00,124.95
-2022-01-08 00:00:00,1367.401
-2022-01-08 01:00:00,126.887
-2022-01-08 02:00:00,109.086
-2022-01-08 03:00:00,106.803
-2022-01-08 04:00:00,95.341
-2022-01-08 05:00:00,101.892
-2022-01-08 06:00:00,116.698
-2022-01-08 07:00:00,134.789
-2022-01-08 08:00:00,122.28
-2022-01-08 09:00:00,112.03
-2022-01-08 10:00:00,107.235
-2022-01-08 11:00:00,107.317
-2022-01-08 12:00:00,117.127
-2022-01-08 13:00:00,1357.485
-2022-01-08 14:00:00,124.043
-2022-01-08 15:00:00,106.754
-2022-01-08 16:00:00,108.607
-2022-01-08 17:00:00,102.222
-2022-01-08 18:00:00,112.257
-2022-01-08 19:00:00,126.2
-2022-01-08 20:00:00,140.582
-2022-01-08 21:00:00,139.268
-2022-01-08 22:00:00,117.773
-2022-01-08 23:00:00,113.515
-2022-01-09 00:00:00,111.217
-2022-01-09 01:00:00,124.314
-2022-01-09 02:00:00,1426.753
-2022-01-09 03:00:00,133.468
-2022-01-09 04:00:00,120.783
-2022-01-09 05:00:00,115.584
-2022-01-09 06:00:00,106.49
-2022-01-09 07:00:00,111.2
-2022-01-09 08:00:00,128.537
-2022-01-09 09:00:00,141.139
-2022-01-09 10:00:00,148.031
-2022-01-09 11:00:00,121.057
-2022-01-09 12:00:00,117.165
-2022-01-09 13:00:00,116.659
-2022-01-09 14:00:00,131.695
-2022-01-09 15:00:00,1491.807
-2022-01-09 16:00:00,132.233
-2022-01-09 17:00:00,124.345
-2022-01-09 18:00:00,122.803
-2022-01-09 19:00:00,111.194
-2022-01-09 20:00:00,115.915
-2022-01-09 21:00:00,126.031
-2022-01-09 22:00:00,138.083
-2022-01-09 23:00:00,139.614
-2022-01-10 00:00:00,124.594
-2022-01-10 01:00:00,119.943
-2022-01-10 02:00:00,120.824
-2022-01-10 03:00:00,143.264
-2022-01-10 04:00:00,1518.844
-2022-01-10 05:00:00,132.923
-2022-01-10 06:00:00,116.26
-2022-01-10 07:00:00,121.806
-2022-01-10 08:00:00,115.909
-2022-01-10 09:00:00,118.556
-2022-01-10 10:00:00,130.719
-2022-01-10 11:00:00,143.146
-2022-01-10 12:00:00,146.257
-2022-01-10 13:00:00,134.042
-2022-01-10 14:00:00,128.8
-2022-01-10 15:00:00,122.119
-2022-01-10 16:00:00,137.027
-2022-01-10 17:00:00,1547.564
-2022-01-10 18:00:00,143.148
-2022-01-10 19:00:00,117.505
-2022-01-10 20:00:00,118.823
-2022-01-10 21:00:00,111.147
-2022-01-10 22:00:00,122.745
-2022-01-10 23:00:00,131.945
-2022-01-11 00:00:00,144.042
-2022-01-11 01:00:00,144.107
-2022-01-11 02:00:00,129.099
-2022-01-11 03:00:00,124.058
-2022-01-11 04:00:00,127.928
-2022-01-11 05:00:00,133.633
-2022-01-11 06:00:00,1548.18
-2022-01-11 07:00:00,137.39
-2022-01-11 08:00:00,121.125
-2022-01-11 09:00:00,126.201
-2022-01-11 10:00:00,118.33
-2022-01-11 11:00:00,121.946
-2022-01-11 12:00:00,129.054
-2022-01-11 13:00:00,148.563
-2022-01-11 14:00:00,142.166
-2022-01-11 15:00:00,132.766
-2022-01-11 16:00:00,126.405
-2022-01-11 17:00:00,126.014
-2022-01-11 18:00:00,139.605
-2022-01-11 19:00:00,1569.566
-2022-01-11 20:00:00,139.269
-2022-01-11 21:00:00,128.056
-2022-01-11 22:00:00,134.599
-2022-01-11 23:00:00,119.596
-2022-01-12 00:00:00,120.469
-2022-01-12 01:00:00,136.697
-2022-01-12 02:00:00,157.967
-2022-01-12 03:00:00,156.392
-2022-01-12 04:00:00,133.435
-2022-01-12 05:00:00,130.321
-2022-01-12 06:00:00,131.799
-2022-01-12 07:00:00,143.919
-2022-01-12 08:00:00,1632.52
-2022-01-12 09:00:00,153.119
-2022-01-12 10:00:00,131.25
-2022-01-12 11:00:00,132.54
-2022-01-12 12:00:00,120.392
-2022-01-12 13:00:00,126.496
-2022-01-12 14:00:00,146.639
-2022-01-12 15:00:00,153.375
-2022-01-12 16:00:00,151.757
-2022-01-12 17:00:00,133.235
-2022-01-12 18:00:00,129.229
-2022-01-12 19:00:00,124.074
-2022-01-12 20:00:00,136.675
-2022-01-12 21:00:00,1638.781
-2022-01-12 22:00:00,143.097
-2022-01-12 23:00:00,127.774
-2022-01-13 00:00:00,127.348
-2022-01-13 01:00:00,118.414
-2022-01-13 02:00:00,125.516
-2022-01-13 03:00:00,138.913
-2022-01-13 04:00:00,159.638
-2022-01-13 05:00:00,167.717
-2022-01-13 06:00:00,137.474
-2022-01-13 07:00:00,132.871
-2022-01-13 08:00:00,134.591
-2022-01-13 09:00:00,147.39
-2022-01-13 10:00:00,1660.744
-2022-01-13 11:00:00,153.915
-2022-01-13 12:00:00,138.446
-2022-01-13 13:00:00,138.362
-2022-01-13 14:00:00,124.885
-2022-01-13 15:00:00,135.014
-2022-01-13 16:00:00,147.23
-2022-01-13 17:00:00,160.939
-2022-01-13 18:00:00,162.998
-2022-01-13 19:00:00,144.095
-2022-01-13 20:00:00,143.399
-2022-01-13 21:00:00,147.318
-2022-01-13 22:00:00,155.809
-2022-01-13 23:00:00,1752.41
-2022-01-14 00:00:00,162.245
-2022-01-14 01:00:00,135.873
-2022-01-14 02:00:00,138.495
-2022-01-14 03:00:00,130.325
-2022-01-14 04:00:00,136.525
-2022-01-14 05:00:00,147.624
-2022-01-14 06:00:00,169.108
-2022-01-14 07:00:00,164.925
-2022-01-14 08:00:00,152.353
-2022-01-14 09:00:00,152.153
-2022-01-14 10:00:00,146.425
-2022-01-14 11:00:00,160.993
-2022-01-14 12:00:00,1797.045
-2022-01-14 13:00:00,159.602
-2022-01-14 14:00:00,138.682
-2022-01-14 15:00:00,144.143
-2022-01-14 16:00:00,133.418
-2022-01-14 17:00:00,146.249
-2022-01-14 18:00:00,158.591
-2022-01-14 19:00:00,174.995
-2022-01-14 20:00:00,174.745
-2022-01-14 21:00:00,156.709
-2022-01-14 22:00:00,147.334
-2022-01-14 23:00:00,139.397
-2022-01-15 00:00:00,154.318
-2022-01-15 01:00:00,1828.184
-2022-01-15 02:00:00,161.387
-2022-01-15 03:00:00,138.081
-2022-01-15 04:00:00,145.169
-2022-01-15 05:00:00,137.504
-2022-01-15 06:00:00,144.891
-2022-01-15 07:00:00,157.377
-2022-01-15 08:00:00,179.42
-2022-01-15 09:00:00,173.707
-2022-01-15 10:00:00,154.874
-2022-01-15 11:00:00,147.523
-2022-01-15 12:00:00,142.095
-2022-01-15 13:00:00,154.387
-2022-01-15 14:00:00,1836.415
-2022-01-15 15:00:00,172.903
-2022-01-15 16:00:00,155.75
-2022-01-15 17:00:00,152.347
-2022-01-15 18:00:00,137.295
-2022-01-15 19:00:00,151.232
-2022-01-15 20:00:00,165.771
-2022-01-15 21:00:00,172.816
-2022-01-15 22:00:00,179.476
-2022-01-15 23:00:00,158.627
-2022-01-16 00:00:00,156.631
-2022-01-16 01:00:00,155.015
-2022-01-16 02:00:00,169.514
-2022-01-16 03:00:00,1927.376
-2022-01-16 04:00:00,171.379
-2022-01-16 05:00:00,146.187
-2022-01-16 06:00:00,151.068
-2022-01-16 07:00:00,136.46
-2022-01-16 08:00:00,148.784
-2022-01-16 09:00:00,159.457
-2022-01-16 10:00:00,176.898
-2022-01-16 11:00:00,180.673
-2022-01-16 12:00:00,153.202
-2022-01-16 13:00:00,146.64
-2022-01-16 14:00:00,142.367
-2022-01-16 15:00:00,156.713
-2022-01-16 16:00:00,1869.826
-2022-01-16 17:00:00,159.356
-2022-01-16 18:00:00,139.482
-2022-01-16 19:00:00,146.594
-2022-01-16 20:00:00,138.353
-2022-01-16 21:00:00,147.956
-2022-01-16 22:00:00,161.561
-2022-01-16 23:00:00,179.351
-2022-01-17 00:00:00,177.391
-2022-01-17 01:00:00,162.681
-2022-01-17 02:00:00,156.051
-2022-01-17 03:00:00,153.318
-2022-01-17 04:00:00,167.785
-2022-01-17 05:00:00,1889.879
-2022-01-17 06:00:00,175.518
-2022-01-17 07:00:00,152.63
-2022-01-17 08:00:00,150.919
-2022-01-17 09:00:00,138.322
-2022-01-17 10:00:00,147.491
-2022-01-17 11:00:00,160.056
-2022-01-17 12:00:00,178.656
-2022-01-17 13:00:00,181.673
-2022-01-17 14:00:00,161.942
-2022-01-17 15:00:00,155.337
-2022-01-17 16:00:00,156.001
-2022-01-17 17:00:00,172.427
-2022-01-17 18:00:00,1930.971
-2022-01-17 19:00:00,175.35
-2022-01-17 20:00:00,157.548
-2022-01-17 21:00:00,149.42
-2022-01-17 22:00:00,138.155
-2022-01-17 23:00:00,153.824
-2022-01-18 00:00:00,164.022
-2022-01-18 01:00:00,178.541
-2022-01-18 02:00:00,175.913
-2022-01-18 03:00:00,163.182
-2022-01-18 04:00:00,155.596
-2022-01-18 05:00:00,156.598
-2022-01-18 06:00:00,174.92
-2022-01-18 07:00:00,1943.069
-2022-01-18 08:00:00,175.514
-2022-01-18 09:00:00,153.555
-2022-01-18 10:00:00,159.857
-2022-01-18 11:00:00,140.548
-2022-01-18 12:00:00,151.645
-2022-01-18 13:00:00,171.133
-2022-01-18 14:00:00,184.087
-2022-01-18 15:00:00,186.022
-2022-01-18 16:00:00,169.222
-2022-01-18 17:00:00,160.682
-2022-01-18 18:00:00,155.899
-2022-01-18 19:00:00,175.664
-2022-01-18 20:00:00,1983.829
-2022-01-18 21:00:00,166.165
-2022-01-18 22:00:00,154.226
-2022-01-18 23:00:00,156.997
-2022-01-19 00:00:00,138.082
-2022-01-19 01:00:00,153.221
-2022-01-19 02:00:00,166.169
-2022-01-19 03:00:00,184.807
-2022-01-19 04:00:00,186.844
-2022-01-19 05:00:00,160.475
-2022-01-19 06:00:00,159.328
-2022-01-19 07:00:00,156.604
-2022-01-19 08:00:00,170.779
-2022-01-19 09:00:00,1953.699
-2022-01-19 10:00:00,174.337
-2022-01-19 11:00:00,159.772
-2022-01-19 12:00:00,155.545
-2022-01-19 13:00:00,144.038
-2022-01-19 14:00:00,154.284
-2022-01-19 15:00:00,170.559
-2022-01-19 16:00:00,183.582
-2022-01-19 17:00:00,188.407
-2022-01-19 18:00:00,167.18
-2022-01-19 19:00:00,159.36
-2022-01-19 20:00:00,156.666
-2022-01-19 21:00:00,173.557
-2022-01-19 22:00:00,1987.288
-2022-01-19 23:00:00,177.865
-2022-01-20 00:00:00,163.114
-2022-01-20 01:00:00,156.693
-2022-01-20 02:00:00,144.524
-2022-01-20 03:00:00,152.626
-2022-01-20 04:00:00,168.239
-2022-01-20 05:00:00,184.592
-2022-01-20 06:00:00,179.993
-2022-01-20 07:00:00,160.76
-2022-01-20 08:00:00,150.605
-2022-01-20 09:00:00,152.302
-2022-01-20 10:00:00,168.073
-2022-01-20 11:00:00,1959.387
-2022-01-20 12:00:00,169.01
-2022-01-20 13:00:00,138.457
-2022-01-20 14:00:00,134.157
-2022-01-20 15:00:00,125.161
-2022-01-20 16:00:00,131.317
-2022-01-20 17:00:00,147.22
-2022-01-20 18:00:00,157.083
-2022-01-20 19:00:00,161.543
-2022-01-20 20:00:00,137.175
-2022-01-20 21:00:00,138.952
-2022-01-20 22:00:00,136.236
-2022-01-20 23:00:00,164.564
-2022-01-21 00:00:00,1740.874
-2022-01-21 01:00:00,169.541
-2022-01-21 02:00:00,149.732
-2022-01-21 03:00:00,142.659
-2022-01-21 04:00:00,125.288
-2022-01-21 05:00:00,141.697
-2022-01-21 06:00:00,163.128
-2022-01-21 07:00:00,177.165
-2022-01-21 08:00:00,176.655
-2022-01-21 09:00:00,148.406
-2022-01-21 10:00:00,132.129
-2022-01-21 11:00:00,135.919
-2022-01-21 12:00:00,165.3
-2022-01-21 13:00:00,1827.618
-2022-01-21 14:00:00,166.309
-2022-01-21 15:00:00,135.738
-2022-01-21 16:00:00,133.862
-2022-01-21 17:00:00,123.65
-2022-01-21 18:00:00,135.488
-2022-01-21 19:00:00,155.016
-2022-01-21 20:00:00,173.728
-2022-01-21 21:00:00,169.987
-2022-01-21 22:00:00,141.456
-2022-01-21 23:00:00,128.32
-2022-01-22 00:00:00,123.628
-2022-01-22 01:00:00,135.527
-2022-01-22 02:00:00,1722.711
-2022-01-22 03:00:00,129.542
-2022-01-22 04:00:00,115.247
-2022-01-22 05:00:00,105.084
-2022-01-22 06:00:00,94.674
-2022-01-22 07:00:00,114.96
-2022-01-22 08:00:00,131.167
-2022-01-22 09:00:00,158.331
-2022-01-22 10:00:00,151.36
-2022-01-22 11:00:00,127.184
-2022-01-22 12:00:00,121.741
-2022-01-22 13:00:00,127.974
-2022-01-22 14:00:00,133.974
-2022-01-22 15:00:00,1511.237
-2022-01-22 16:00:00,137.055
-2022-01-22 17:00:00,122.562
-2022-01-22 18:00:00,128.606
-2022-01-22 19:00:00,110.959
-2022-01-22 20:00:00,118.029
-2022-01-22 21:00:00,137.026
-2022-01-22 22:00:00,151.952
-2022-01-22 23:00:00,149.701
-2022-01-23 00:00:00,132.785
-2022-01-23 01:00:00,121.075
-2022-01-23 02:00:00,120.445
-2022-01-23 03:00:00,141.155
-2022-01-23 04:00:00,1571.349
-2022-01-23 05:00:00,153.85
-2022-01-23 06:00:00,140.102
-2022-01-23 07:00:00,132.664
-2022-01-23 08:00:00,106.751
-2022-01-23 09:00:00,117.672
-2022-01-23 10:00:00,136.577
-2022-01-23 11:00:00,149.776
-2022-01-23 12:00:00,148.995
-2022-01-23 13:00:00,126.983
-2022-01-23 14:00:00,112.3
-2022-01-23 15:00:00,118.837
-2022-01-23 16:00:00,124.629
-2022-01-23 17:00:00,1569.137
-2022-01-23 18:00:00,130.232
-2022-01-23 19:00:00,122.417
-2022-01-23 20:00:00,106.412
-2022-01-23 21:00:00,88.646
-2022-01-23 22:00:00,104.498
-2022-01-23 23:00:00,126.28
-2022-01-24 00:00:00,140.283
-2022-01-24 01:00:00,135.156
-2022-01-24 02:00:00,118.653
-2022-01-24 03:00:00,98.469
-2022-01-24 04:00:00,90.263
-2022-01-24 05:00:00,91.658
-2022-01-24 06:00:00,1352.967
-2022-01-24 07:00:00,113.495
-2022-01-24 08:00:00,92.416
-2022-01-24 09:00:00,72.84
-2022-01-24 10:00:00,71.41
-2022-01-24 11:00:00,82.51
-2022-01-24 12:00:00,115.772
-2022-01-24 13:00:00,135.958

wgz_forecast(lh)/load/load_inference.py

@@ -1,25 +0,0 @@
-import joblib
-import numpy as np
-
-
-def load_load_model(path):
-    gbm = joblib.load(path)
-    return gbm
-
-def load_forecast(inputs: np.ndarray, model):
-    """_summary_
-
-    Args:
-        inputs (np.ndarray): 输入序列
-        model (_type_): _description_
-    """
-    out = model.predict([inputs])
-    return out
-
-
-if __name__ == '__main__':
-    model = load_load_model('./models/load_pred.joblib')
-    inputs = np.random.randn(24)
-    print(inputs.shape)
-    out = load_forecast(inputs, model)
-    print(out)

wgz_forecast(lh)/load/load_train.py

@@ -1,73 +0,0 @@
-import pandas as pd
-import numpy as np
-import xgboost as xgb
-from sklearn.model_selection import train_test_split
-from sklearn.metrics import mean_squared_error, r2_score
-import joblib
-from logzero import logger
-
-def time_series_to_supervised(data, columns, n_in=24, n_out=1,dropnan=True):
-    """
-    :param data:作为列表或2D NumPy数组的观察序列。需要。
-    :param n_in:作为输入的滞后观察数（X）。值可以在[1..len（数据）]之间可选。默认为1。
-    :param n_out:作为输出的观测数量（y）。值可以在[0..len（数据）]之间。可选的。默认为1。
-    :param dropnan:Boolean是否删除具有NaN值的行。可选的。默认为True。
-    :return:
-    """
-    logger.info(f"正在处理训练数据：size：{data.shape}")
-    n_vars = 1 if type(data) is list else data.shape[1]
-    df = pd.DataFrame(data)
-    origNames = columns
-    cols, names = list(), list()
-    cols.append(df.shift(0))
-    names += [('%s' % origNames[j]) for j in range(n_vars)]
-    n_in = max(1, n_in)
-    for i in range(n_in-1, 0, -1):
-        time = '(t-%d)' % i
-        cols.append(df.shift(i))
-        names += [('%s%s' % (origNames[j], time)) for j in range(n_vars)]
-    n_out = max(n_out, 0)
-    for i in range(1, n_out+1):
-        time = '(t+%d)' % i
-        cols.append(df.shift(-i))
-        names += [('%s%s' % (origNames[j], time)) for j in range(n_vars)]
-    agg = pd.concat(cols, axis=1)
-    agg.columns = names
-    if dropnan:
-        agg.dropna(inplace=True)
-    return agg
-
-def train_model(train_data: pd.DataFrame):
-    """训练模型的函数，需要根据模型类型实际调整
-
-    Args:
-        data (pd.DataFrame): 训练集
-    """
-    # 特征和输出列名，需要根据业务场景灵活处理
-    fea_cols  = train_data.columns[:-1].tolist()
-    out_cols = train_data.columns[-1:].tolist()
-    logger.info(fea_cols, out_cols)
-    X = train_data[fea_cols]
-    y = train_data[out_cols]
-    train_X,test_X,train_y,test_y = train_test_split(X, y, test_size=0.2, random_state=42)
-    valid_X,test_X,valid_y,test_y = train_test_split(test_X, test_y, test_size=0.5, random_state=42)
-
-    # 参数
-    other_params = {'learning_rate': 0.1, 'n_estimators': 150, 'max_depth': 10, 'min_child_weight': 1, 'seed': 0, 'subsample': 0.8, 'colsample_bytree': 0.8, 'gamma': 0, 'reg_alpha': 0, 'reg_lambda': 1}
-    print(train_X.shape, train_y.shape)
-
-    gbm = xgb.XGBRegressor(objective='reg:squarederror',**other_params)
-
-    gbm.fit(train_X.values, train_y.values, eval_set=[(valid_X.values, valid_y.values)], early_stopping_rounds=20)
-
-    y_pred = gbm.predict(test_X.values)
-
-    logger.info(f"Root Mean Squared Error on Test set: {np.sqrt(mean_squared_error(test_y, y_pred))}")
-    logger.info(f"R2 score on Test set: {r2_score(test_y, y_pred)}")
-    joblib.dump(gbm, './models/load_pred.joblib')
-    logger.info(f"save_path: ./models/load_pred.joblib")
-
-if __name__ == '__main__':
-    data = pd.read_csv('./data/load_data_hourly.csv', index_col=0)
-    agg = time_series_to_supervised(data.values, data.columns, 24, 1)
-    train_model(agg)

wgz_forecast(lh)/pv/pv_inference.py

@@ -1,26 +0,0 @@
-import joblib
-import numpy as np
-
-
-def load_model(path):
-    gbm = joblib.load(path)
-    return gbm
-
-def pv_forecast(inputs: np.ndarray, model):
-    """_summary_
-
-    Args:
-        inputs (np.ndarray): 输入序列
-        model (_type_): _description_
-    """
-    out = model.predict([inputs])
-    return out
-
-
-# if __name__ == '__main__':
-#     model = load_model('models/pv_pred.joblib')
-#     inputs = np.random.randn(24)
-#
-#     print(inputs.shape)
-#     out = pv_forecast(inputs, model)
-#     print(out)

wgz_forecast(lh)/pv/pv_train.py

@@ -1,79 +0,0 @@
-import pandas as pd
-import numpy as np
-import xgboost as xgb
-from sklearn.model_selection import train_test_split
-from sklearn.metrics import mean_squared_error, r2_score
-import joblib
-from logzero import logger
-
-def time_series_to_supervised(data, columns, n_in=24, n_out=1,dropnan=True):
-    """
-    :param data:作为列表或2D NumPy数组的观察序列。需要。
-    :param n_in:作为输入的滞后观察数（X）。值可以在[1..len（数据）]之间可选。默认为1。
-    :param n_out:作为输出的观测数量（y）。值可以在[0..len（数据）]之间。可选的。默认为1。
-    :param dropnan:Boolean是否删除具有NaN值的行。可选的。默认为True。
-    :return:
-    """
-    logger.info(f"正在处理训练数据：size：{data.shape}")
-    n_vars = 1 if type(data) is list else data.shape[1]
-    df = pd.DataFrame(data)
-    origNames = columns
-    cols, names = list(), list()
-    cols.append(df.shift(0))
-    names += [('%s' % origNames[j]) for j in range(n_vars)]
-    n_in = max(1, n_in)
-    for i in range(n_in-1, 0, -1):
-        time = '(t-%d)' % i
-        cols.append(df.shift(i))
-        names += [('%s%s' % (origNames[j], time)) for j in range(n_vars)]
-    n_out = max(n_out, 0)
-    for i in range(1, n_out+1):
-        time = '(t+%d)' % i
-        cols.append(df.shift(-i))
-        names += [('%s%s' % (origNames[j], time)) for j in range(n_vars)]
-    agg = pd.concat(cols, axis=1)
-    agg.columns = names
-    if dropnan:
-        agg.dropna(inplace=True)
-    return agg
-
-def train_model(train_data: pd.DataFrame):
-    """训练模型的函数，需要根据模型类型实际调整
-
-    Args:
-        data (pd.DataFrame): 训练集
-    """
-    # 特征和输出列名，需要根据业务场景灵活处理
-    fea_cols  = train_data.columns[:-1].tolist()
-    out_cols = train_data.columns[-1:].tolist()
-    logger.info(fea_cols, out_cols)
-    X = train_data[fea_cols]
-    y = train_data[out_cols]
-    # 8:1:1
-    train_X,test_X,train_y,test_y = train_test_split(X, y, test_size=0.2, random_state=42)
-
-    valid_X,test_X,valid_y,test_y = train_test_split(test_X, test_y, test_size=0.5, random_state=42)
-
-    # 参数
-    other_params = {'learning_rate': 0.1, 'n_estimators': 150, 'max_depth': 10, 'min_child_weight': 1, 'seed': 0, 'subsample': 0.8, 'colsample_bytree': 0.8, 'gamma': 0, 'reg_alpha': 0, 'reg_lambda': 1}
-    print(train_X.shape, train_y.shape)
-    print("测试集test_X:",test_X.shape, "测试集test_y:",test_y.shape)
-
-    gbm = xgb.XGBRegressor(objective='reg:squarederror',**other_params)
-
-    gbm.fit(train_X.values,
-            train_y.values,
-            eval_set=[(valid_X.values, valid_y.values)],
-            early_stopping_rounds=20)
-
-    y_pred = gbm.predict(test_X.values)
-
-    logger.info(f"Root Mean Squared Error on Test set: {np.sqrt(mean_squared_error(test_y, y_pred))}")
-    logger.info(f"R2 score on Test set: {r2_score(test_y, y_pred)}")
-    joblib.dump(gbm, './models/pv_pred.joblib')
-    logger.info(f"save_path: ./models/pv_pred.joblib")
-
-if __name__ == '__main__':
-    data = pd.read_csv('./data/pv_data_hourly.csv', index_col=0)
-    agg = time_series_to_supervised(data.values, data.columns, 24, 1)
-    train_model(agg)

wgz_forecast(lh)/requirements.txt

@@ -1,13 +0,0 @@
-EMD_signal==1.6.0
-matplotlib==3.7.0
-numpy==1.22.0
-pandas==1.5.3
-PyEMD==1.0.0
-scikit_learn==1.2.1
-seaborn==0.13.2
-xlrd==2.0.1
-vmdpy==0.2
-scikit-optimize==0.10.1
-xgboost==1.6.0
-logzero==1.7.0
-flask==3.1.0

wgz_forecast(lh)/run.py

@@ -1,103 +0,0 @@
-from flask import request, Flask, jsonify
-from pv.pv_inference import load_model, pv_forecast
-from carbon.carbon_inference import load_carbon_model,carbon_forecast
-from water.water_inference import load_water_model,water_forecast
-from load.load_inference import load_load_model,load_forecast
-from logzero import logger
-import numpy as np
-app = Flask(__name__)
-pv_model_gbm = load_model('./pv/models/pv_pred.joblib')
-carbon_model_gbm = load_carbon_model('./carbon/models/carbon_pred.joblib')
-water_model_gbm = load_water_model('./water/models/water_pred.joblib')
-load_model_gbm = load_load_model('./load/models/load_pred.joblib')
-
-
-@app.route('/pv', methods=['POST'])
-def run_pv_forecast():
-    try:
-        data = request.json
-        if not data or 'inputs' not in data:
-            return jsonify({"error": "Invalid data"}), 400
-        inputs = data.get('inputs')
-        # 转为1维数组
-        inputs = np.array(inputs)
-        if inputs.ndim != 1:
-            inputs = inputs.flatten()
-        logger.info(f"pv history inputs: {inputs}")
-        # 调用预测函数
-        out = pv_forecast(inputs, pv_model_gbm)
-        # 将 NumPy 数组转换为列表
-        results = {"result": out.tolist()}  # 使用 .tolist() 转换为列表
-        return jsonify(results), 200
-    except Exception as e:
-        logger.error(f"Error in pv_forecast: {e}", exc_info=True)
-        return jsonify({"error": str(e)}), 500
-    
-@app.route('/carbon', methods=['POST'])
-def run_carbon_forecast():
-
-    try:
-        data = request.json
-        if not data or 'inputs' not in data:
-            return jsonify({"error": "Invalid data"}), 400
-        inputs = data.get('inputs')
-        # 转为1维数组
-        inputs = np.array(inputs)
-        if inputs.ndim != 1:
-            inputs = inputs.flatten()
-        logger.info(f"carbon history inputs: {inputs}")
-        # 调用预测函数
-        out = carbon_forecast(inputs, carbon_model_gbm)
-        # 将 NumPy 数组转换为列表
-        results = {"result": out.tolist()}  # 使用 .tolist() 转换为列表
-        return jsonify(results), 200
-    except Exception as e:
-        logger.error(f"Error in carbon_forecast: {e}", exc_info=True)
-        return jsonify({"error": str(e)}), 500
-
-@app.route('/water', methods=['POST'])
-def run_water_forecast():
-
-    try:
-        data = request.json
-        if not data or 'inputs' not in data:
-            return jsonify({"error": "Invalid data"}), 400
-        inputs = data.get('inputs')
-        # 转为1维数组
-        inputs = np.array(inputs)
-        if inputs.ndim != 1:
-            inputs = inputs.flatten()
-        logger.info(f"water history inputs: {inputs}")
-        # 调用预测函数
-        out = water_forecast(inputs, water_model_gbm)
-        # 将 NumPy 数组转换为列表
-        results = {"result": out.tolist()}  # 使用 .tolist() 转换为列表
-        return jsonify(results), 200
-    except Exception as e:
-        logger.error(f"Error in water_forecast: {e}", exc_info=True)
-        return jsonify({"error": str(e)}), 500
-@app.route('/load', methods=['POST'])
-def run_load_forecast():
-
-    try:
-        data = request.json
-        if not data or 'inputs' not in data:
-            return jsonify({"error": "Invalid data"}), 400
-        inputs = data.get('inputs')
-        # 转为1维数组
-        inputs = np.array(inputs)
-        if inputs.ndim != 1:
-            inputs = inputs.flatten()
-        logger.info(f"load history inputs: {inputs}")
-        # 调用预测函数
-        out = load_forecast(inputs, load_model_gbm)
-        # 将 NumPy 数组转换为列表
-        results = {"result": out.tolist()}  # 使用 .tolist() 转换为列表
-        return jsonify(results), 200
-    except Exception as e:
-        logger.error(f"Error in load_forecast: {e}", exc_info=True)
-        return jsonify({"error": str(e)}), 500
-
-if __name__=='__main__':
-    app.run(host='0.0.0.0', port='2467')
-

wgz_forecast(lh)/test.py

@@ -1,21 +0,0 @@
-import requests
-import pandas as pd
-import numpy as np
-# 这里使用的是load进行预测，可改(load/pv/water/carbon)
-url = "http://127.0.0.1:2467/load"
-data = {
-    "inputs":(np.random.randn(24)*100).tolist()
-}
-
-response = requests.post(url, json=data)
-
-print("Status Code:", response.status_code)  # 打印状态码
-print("Response Content:", response.text)    # 打印响应内容
-
-if response.status_code == 200:
-    try:
-        print(response.json())  # 尝试解析 JSON
-    except requests.exceptions.JSONDecodeError as e:
-        print("JSON Decode Error:", e)
-else:
-    print("Error:", response.status_code)

wgz_forecast(lh)/water/water_inference.py

@@ -1,25 +0,0 @@
-import joblib
-import numpy as np
-
-
-def load_water_model(path):
-    gbm = joblib.load(path)
-    return gbm
-
-def water_forecast(inputs: np.ndarray, model):
-    """_summary_
-
-    Args:
-        inputs (np.ndarray): 输入序列
-        model (_type_): _description_
-    """
-    out = model.predict([inputs])
-    return out
-
-
-if __name__ == '__main__':
-    model = load_water_model('./models/water_pred.joblib')
-    inputs = np.random.randn(24)
-    print(inputs.shape)
-    out = water_forecast(inputs, model)
-    print(out)

wgz_forecast(lh)/water/water_train.py

@@ -1,73 +0,0 @@
-import pandas as pd
-import numpy as np
-import xgboost as xgb
-from sklearn.model_selection import train_test_split
-from sklearn.metrics import mean_squared_error, r2_score
-import joblib
-from logzero import logger
-
-def time_series_to_supervised(data, columns, n_in=24, n_out=1,dropnan=True):
-    """
-    :param data:作为列表或2D NumPy数组的观察序列。需要。
-    :param n_in:作为输入的滞后观察数（X）。值可以在[1..len（数据）]之间可选。默认为1。
-    :param n_out:作为输出的观测数量（y）。值可以在[0..len（数据）]之间。可选的。默认为1。
-    :param dropnan:Boolean是否删除具有NaN值的行。可选的。默认为True。
-    :return:
-    """
-    logger.info(f"正在处理训练数据：size：{data.shape}")
-    n_vars = 1 if type(data) is list else data.shape[1]
-    df = pd.DataFrame(data)
-    origNames = columns
-    cols, names = list(), list()
-    cols.append(df.shift(0))
-    names += [('%s' % origNames[j]) for j in range(n_vars)]
-    n_in = max(1, n_in)
-    for i in range(n_in-1, 0, -1):
-        time = '(t-%d)' % i
-        cols.append(df.shift(i))
-        names += [('%s%s' % (origNames[j], time)) for j in range(n_vars)]
-    n_out = max(n_out, 0)
-    for i in range(1, n_out+1):
-        time = '(t+%d)' % i
-        cols.append(df.shift(-i))
-        names += [('%s%s' % (origNames[j], time)) for j in range(n_vars)]
-    agg = pd.concat(cols, axis=1)
-    agg.columns = names
-    if dropnan:
-        agg.dropna(inplace=True)
-    return agg
-
-def train_model(train_data: pd.DataFrame):
-    """训练模型的函数，需要根据模型类型实际调整
-
-    Args:
-        data (pd.DataFrame): 训练集
-    """
-    # 特征和输出列名，需要根据业务场景灵活处理
-    fea_cols  = train_data.columns[:-1].tolist()
-    out_cols = train_data.columns[-1:].tolist()
-    logger.info(fea_cols, out_cols)
-    X = train_data[fea_cols]
-    y = train_data[out_cols]
-    train_X,test_X,train_y,test_y = train_test_split(X, y, test_size=0.2, random_state=42)
-    valid_X,test_X,valid_y,test_y = train_test_split(test_X, test_y, test_size=0.5, random_state=42)
-
-    # 参数
-    other_params = {'learning_rate': 0.1, 'n_estimators': 150, 'max_depth': 10, 'min_child_weight': 1, 'seed': 0, 'subsample': 0.8, 'colsample_bytree': 0.8, 'gamma': 0, 'reg_alpha': 0, 'reg_lambda': 1}
-    print(train_X.shape, train_y.shape)
-
-    gbm = xgb.XGBRegressor(objective='reg:squarederror',**other_params)
-
-    gbm.fit(train_X.values, train_y.values, eval_set=[(valid_X.values, valid_y.values)], early_stopping_rounds=20)
-
-    y_pred = gbm.predict(test_X.values)
-
-    logger.info(f"Root Mean Squared Error on Test set: {np.sqrt(mean_squared_error(test_y, y_pred))}")
-    logger.info(f"R2 score on Test set: {r2_score(test_y, y_pred)}")
-    joblib.dump(gbm, './models/water_pred.joblib')
-    logger.info(f"save_path: ./models/water_pred.joblib")
-
-if __name__ == '__main__':
-    data = pd.read_csv('./data/preprocessed_example.csv', index_col=0)
-    agg = time_series_to_supervised(data.values, data.columns, 24, 1)
-    train_model(agg)