Datawhale AI夏令营task02

主要目的：上分

学习内容：

通过时间序列挖掘并进行可视化分析，对有价值的信息进行分析提取，有助于对题目的理解。

代码：

import pandas as pd  
from sklearn.linear_model import LinearRegression  
import numpy as np  
from pathlib import Path  

# 设置基础路径  
base_path = Path("data")  

# 读取市场数据  
electricity_price = pd.read_csv(base_path / "electricity price.csv")  
# 读取单位数据  
unit = pd.read_csv(base_path / "unit.csv")  

# 检查并处理缺失值  
# 如果存在缺失值，先用均值填充  
if electricity_price["clearing price (CNY/MWh)"].isna().any():  
    mean_value = electricity_price["clearing price (CNY/MWh)"].mean()  
    electricity_price["clearing price (CNY/MWh)"].fillna(mean_value, inplace=True)  

# 如果需要，可以选择删除含有缺失值的行  
# electricity_price.dropna(subset=["clearing price (CNY/MWh)"], inplace=True)  

# 准备示例提交数据  
sample_submit = electricity_price[electricity_price["clearing price (CNY/MWh)"].isna()].drop(columns="demand")  
sample_submit.to_csv(base_path / "sample_submit.csv", index=False)  

# 预处理时间戳  
electricity_price["timestamp"] = pd.to_datetime(  
    electricity_price["day"] + " " + electricity_price["time"].str.replace("24:00:00", "00:00"))  

mask = electricity_price['timestamp'].dt.time == pd.Timestamp('00:00:00').time()  
# 对于这些行，增加一天  
electricity_price.loc[mask, 'timestamp'] += pd.Timedelta(days=1)  

# 排序单位数据  
sorted_unit = unit.sort_values(by='Capacity（MW）')  

# 预计算累积容量  
sorted_unit['cumulative_capacity'] = sorted_unit['Capacity（MW）'].cumsum()  

prices = []  
# 找到满足总需求的最后报价单位  
for demand in electricity_price["demand"]:  
    price = sorted_unit[sorted_unit['cumulative_capacity'] >= demand]["coal consumption (g coal/KWh)"].iloc[0]  
    prices.append(price)  

# 创建线性回归模型实例  
model = LinearRegression()  

# 训练模型  
train_length = 55392  
# 将价格列表转换为 NumPy 数组并重塑  
prices = np.array(prices).reshape(-1, 1)  
# 训练特征矩阵 X，由价格数据组成  
X = prices[:train_length]  
# 训练目标变量 y，清算价格  
y = electricity_price["clearing price (CNY/MWh)"].iloc[:train_length].values.reshape(-1, 1)  

# 拟合线性回归模型  
model. Fit(X, y)


y_pred = model.predict(prices[train_length:])
y_pred = y_pred.flatten()  # 2维矩阵转为1维
y_pred[:5]


sample_submit["clearing price (CNY/MWh)"] = y_pred
sample_submit.head()


sample_submit.to_csv("sample_submit1.csv", index=False)