电力需求预测挑战赛

理解赛题

【训练时序预测模型助力电力需求预测

赛题任务

给定多个房屋对应电力消耗历史N天的相关序列数据等信息，预测房屋对应电力的消耗。

赛题数据

赛题数据由训练集和测试集组成，为了保证比赛的公平性，将每日日期进行脱敏，用1-N进行标识。
即1为数据集最近一天，其中1-10为测试集数据。
数据集由字段id（房屋id）、 dt（日标识）、type（房屋类型）、target（实际电力消耗）组成。

注意时间穿越问题，不要用未来的数据预测过去

优化思路

0.特征优化
1.窗口从三个变成7，10个
2.除了均值，还有最值，中位数，方差
3.差分统计，增大减小的趋势
4.K值交叉验证（可能出现穿越问题）
5.模型融合

baseline代码

学习文件的输入与输出

# 1. 导入需要用到的相关库
# 导入 pandas 库，用于数据处理和分析
import pandas as pd
# 导入 numpy 库，用于科学计算和多维数组操作
import numpy as np

# 2. 读取训练集和测试集
# 使用 read_csv() 函数从文件中读取训练集数据，文件名为 'train.csv'
train = pd.read_csv('./data/train.csv')
# 使用 read_csv() 函数从文件中读取测试集数据，文件名为 'train.csv'
test = pd.read_csv('./data/test.csv')

# 3. 计算训练数据最近11-20单位时间内对应id的目标均值
target_mean = train[train['dt']<=20].groupby(['id'])['target'].mean().reset_index()

# 4. 将target_mean作为测试集结果进行合并
test = test.merge(target_mean, on=['id'], how='left')

# 5. 保存结果文件到本地
test[['id','dt','target']].to_csv('task01_submit.csv', index=None)

task02代码-lightgbm

1.数据可视化
2.在代码加入print语句，作为断点，便于调试

#安装lightgbm
 !pip install lightgbm==3.3.0
 import numpy as np
 import pandas as pd
 import lightgbm as lgb
 from sklearn.metrics import mean_squared_log_error, mean_absolute_error, mean_squared_error
 import tqdm
 import sys
 import os
 import gc
 import argparse
 import warnings
 warnings.filterwarnings('ignore')

# 读取训练数据和测试数据
 train = pd.read_csv('./data/data283931/train.csv')
 test = pd.read_csv('./data/data283931/test.csv')

 # 不同type类型对应target的柱状图
 import matplotlib.pyplot as plt
 type_target_df = train.groupby('type')['target'].mean().reset_index()
 plt.figure(figsize=(8, 4))
 plt.bar(type_target_df['type'], type_target_df['target'], color=['blue', 
'green'])
 plt.xlabel('Type')
 plt.ylabel('Average Target Value')
 plt.title('Bar Chart of Target by Type')
 plt.show()

 print("0Look me!!!!!")

# id为00037f39cf的按dt为序列关于target的折线图
 specific_id_df = train[train['id'] == '00037f39cf']
 plt.figure(figsize=(10, 5))
 plt.plot(specific_id_df['dt'], specific_id_df['target'], marker='o', 
linestyle='-')
 plt.xlabel('DateTime')
 plt.ylabel('Target Value')
 plt.title("Line Chart of Target for ID '00037f39cf'")
 plt.show()

 print("1Look me!!!!!")

 # 合并训练数据和测试数据，并进⾏排序
data = pd.concat([test, train], axis=0, ignore_index=True)
 data = data.sort_values(['id','dt'], ascending=False).reset_index(drop=True)

 # 历史平移
 for i in range(10,30):
  data[f'last{i}_target'] = data.groupby(['id'])['target'].shift(i)

 # 窗⼝统计
 data[f'win3_sum_target'] = 0
 for i in range(10,30):
  data[f'win3_sum_target'] += data[f'last{i}_target']
 data[f'win3_mean_target'] = data[f'win3_sum_target'] / 20.0

# data[f'win3_mean_target'] = (data['last10_target'] + data['last11_target'] + data['last12_target']) / 3

 # 进⾏数据切分
 train = data[data.target.notnull()].reset_index(drop=True)
 test = data[data.target.isnull()].reset_index(drop=True)
 # 确定输⼊特征
 train_cols = [f for f in data.columns if f not in ['id','target']]



 print("2Look me!!!!!")

#############
 def time_model(lgb, train_df, test_df, cols): 
 # 训练集和验证集切分
     print("3Look me!!!!!")
     trn_x, trn_y = train_df[train_df.dt>=31][cols], train_df[train_df.dt>=31]['target']
     val_x, val_y = train_df[train_df.dt<=30][cols], train_df[train_df.dt<=30]['target']
     # 构建模型输⼊数据
     print("4Look me!!!!!")
     train_matrix = lgb.Dataset(trn_x, label=trn_y)
     valid_matrix = lgb.Dataset(val_x, label=val_y)
     # lightgbm参数
     print("5Look me!!!!!")
     lgb_params = {
      'boosting_type': 'gbdt',
      'objective': 'regression',
      'metric': 'mse',
      'min_child_weight': 5,
      #'max_depth' : 25,
      'num_leaves': 2 ** 4 + 2 ** 3,
      # 2 ** 5
      'lambda_l2': 10,
      'feature_fraction': 0.8,
      'bagging_fraction': 0.8,
      'bagging_freq': 4,
      'learning_rate': 0.05,
      'seed': 2024,
      'nthread' : 16,
      'verbose' : -1,
      }
     print("5.1Look me!!!!!")
     # 训练模型

     model = lgb.train(lgb_params, train_matrix, 50000, valid_sets=
     [train_matrix, valid_matrix], categorical_feature=[], verbose_eval=500, early_stopping_rounds=500)
     print("5.2Look me!!!!!")
    # 验证集和测试集结果预测
     val_pred = model.predict(val_x, num_iteration=model.best_iteration)
     test_pred = model.predict(test_df[cols], 
     num_iteration=model.best_iteration)
     # 离线分数评估  
     score = mean_squared_error(val_pred, val_y)
     print(score)
     print("6Look me!!!!!")
     return val_pred, test_pred
     print("7Look me!!!!!")
 print("8Look me!!!!!")
 lgb_oof, lgb_test = time_model(lgb, train, test, train_cols)
     


 print("9Look me!!!!!")




 # 保存结果⽂件到本地
 test['target'] = lgb_test
 test[['id','dt','target']].to_csv('task02_submit_5.csv', index=None)