基于pandas和sales模型的电子商务商品销售预测

import pandas as pd

# 加载数据
data = pd.read_csv('sales_data.csv')

# 查看数据结构
print(data.head())

# 处理缺失值
data = data.dropna()

# 处理异常值（可选）
# data = data[(data['sales'] > 0) & (data['price'] > 0)]

# 转换日期格式
data['date'] = pd.to_datetime(data['date'])

# 提取时间特征
data['year'] = data['date'].dt.year
data['month'] = data['date'].dt.month
data['day'] = data['date'].dt.day
data['week_of_year'] = data['date'].dt.isocalendar().week

# 创建时间相关的特征
data['day_of_week'] = data['date'].dt.dayofweek

# 创建滞后特征（例如，前一天的销售量）
data['sales_lag_1'] = data.groupby('product_id')['sales'].shift(1)

# 创建滑动窗口统计特征（例如，过去7天的平均销售量）
data['sales_rolling_mean_7'] = data.groupby('product_id')['sales'].rolling(window=7).mean().reset_index(0, drop=True)

# 删除包含NaN的行（由于滞后特征和滑动窗口统计特征）
data = data.dropna()

from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import mean_squared_error

# 选择特征和目标变量
features = ['year', 'month', 'day', 'week_of_year', 'day_of_week', 'sales_lag_1', 'sales_rolling_mean_7']
target = 'sales'

# 划分训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(data[features], data[target], test_size=0.2, random_state=42)

# 初始化模型
model = RandomForestRegressor(n_estimators=100, random_state=42)

# 训练模型
model.fit(X_train, y_train)

# 预测
y_pred = model.predict(X_test)

# 评估模型
mse = mean_squared_error(y_test, y_pred)
print(f'Mean Squared Error: {mse}')

# 假设你有一个未来的数据集 future_data
future_data = pd.read_csv('future_sales_data.csv')

# 进行相同的预处理和特征工程步骤
# ...

# 使用模型进行预测
future_predictions = model.predict(future_data[features])

# 将预测结果添加到未来数据集中
future_data['predicted_sales'] = future_predictions