使用您提供的 Dataframe ：

import pandas as pd

df = pd.DataFrame(
    {
        "ID": [1, 1, 1, 1, 1, 2, 2, 2],
        "week": [1, 1, 2, 2, 2, 1, 2, 2],
        "date": [
            "23/07/22",
            "24/07/22",
            "29/07/22",
            "30/07/22",
            "01/08/22",
            "22/07/22",
            "29/07/22",
            "02/08/22",
        ],
    }
)

下面是使用Pandas groupby、apply、explode和concat方法执行此操作的一种方法：

# Prep data
df["date"] = pd.to_datetime(df["date"], dayfirst=True)
df = df.sort_values(["week", "date"]).reset_index(drop=True)

# Add day number to df as new column "day_num"
tmp = df.groupby("week").agg(set)
tmp["day_num"] = tmp.apply(lambda x: [i + 1 for i in range(len(x["date"]))], axis=1)
tmp["date"] = tmp["date"].apply(sorted)
tmp = tmp.explode(["date", "day_num"])
df = pd.concat([df.set_index('date'), tmp.set_index('date').drop(columns="ID")], axis=1)

# Evaluate mean as new column "mean_day"
df = (
    df.sort_values(["ID", "week"])
    .groupby(["ID", "week"])
    .agg(pd.DataFrame.mean)
    .reset_index()
    .rename(columns={"date": "mean_day"})
)

然后：

print(df)
# Output
   ID  week  day_num
0   1     1      2.5
1   1     2      2.0
2   2     1      1.0
3   2     2      2.5