R语言如何按分组计算数据框指定行的价格与成交量平均值

下面是两种主流数据分析工具的简便实现方式，你之前用head仅返回第一个分组结果，是因为没有将计算逻辑作用到每个分组的独立上下文里，按下方写法即可得到目标结构的结果：

Python Pandas 实现 #

import pandas as pd

# 假设原始数据框命名为 df
stat_result = df.groupby('Stock').apply(
    lambda group: pd.Series({
        'Pr/3D/Avg': group[group['Day'] <= 3]['Price'].mean(),
        'Pr/5D/Avg': group[group['Day'] <= 5]['Price'].mean(),
        'Pr/20D/Avg': group[(group['Day'] >=5) & (group['Day'] <=25)]['Price'].mean(),
        'Vo/3D/Avg': group[group['Day'] <= 3]['Volume'].mean(),
        'Vo/5D/Avg': group[group['Day'] <= 5]['Volume'].mean(),
        'Vo/20D/Avg': group[(group['Day'] >=5) & (group['Day'] <=25)]['Volume'].mean()
    })
).reset_index()

如果每个股票分组的Day已经按升序排好，还可以用位置索引优化性能：

stat_result = df.groupby('Stock').apply(
    lambda group: pd.Series({
        'Pr/3D/Avg': group['Price'].head(3).mean(),
        'Pr/5D/Avg': group['Price'].head(5).mean(),
        'Pr/20D/Avg': group['Price'].iloc[4:].mean(), # 索引从0开始，第4位对应第5天
        'Vo/3D/Avg': group['Volume'].head(3).mean(),
        'Vo/5D/Avg': group['Volume'].head(5).mean(),
        'Vo/20D/Avg': group['Volume'].iloc[4:].mean()
    })
).reset_index()

R dplyr 实现 #

library(dplyr)

# 假设原始数据框命名为 df
stat_result <- df %>%
  group_by(Stock) %>%
  summarise(
    `Pr/3D/Avg` = mean(Price[Day <= 3]),
    `Pr/5D/Avg` = mean(Price[Day <= 5]),
    `Pr/20D/Avg` = mean(Price[Day >=5 & Day <=25]),
    `Vo/3D/Avg` = mean(Volume[Day <= 3]),
    `Vo/5D/Avg` = mean(Volume[Day <= 5]),
    `Vo/20D/Avg` = mean(Volume[Day >=5 & Day <=25])
  )

内容的提问来源于stack exchange，提问作者kartheek sista