在 python 中分类后，每周和每年对数据进行分组

Question

问题陈述：数据由 csv 中的两列组成 - 日期和产品数据：

  Date                  Prod
  1/2/2018  7:43:00 PM     A
  1/1/2018  11:41:00 AM    B
  1/1/2018  7:57:00 AM     C
  1/2/2018  1:56:00 PM     A
  1/5/2018  3:29:00 AM     A
  1/3/2018  7:23:00 AM     C
  1/3/2018  1:26:00 PM     B
  1/5/2018  2:08:00 AM     A
  1/5/2018  3:47:00 PM     B

我需要返回两个以产品为键的json数据，值应该是产品的频率

1. 每周一次
2. 每年

喜欢：

1. [{"A":{"Week1":"3","Week2":"3","Week3":"5",...},{"B":{"Week1":"5","Week2":"7","Week3":"4",...},{"C":{...}}]

2. [{"A":{"2018":"3","2019":"3","2020":"5",...},{"B":{"2018":"5","2019":"7","2020":"4",...},{"C":{...}}]

我试过了：

df['Date'] = pd.to_datetime(df['Date'])

weekly_series = df.groupby(pd.Grouper(key='Date', freq='W'))['Date'].count()

weekly_series.index = weekly_series.index.week

Answer 1

准备你的数据框：

df["Date"] = pd.to_datetime(df["Date"])
df["Year"] = df["Date"].dt.isocalendar().year
df["Week"] = df["Date"].dt.isocalendar().week

创建您的字典：

dyear = df.groupby("Prod")[["Year"]] \
          .apply(lambda x: x.value_counts("Year").to_dict()).to_dict()

dweek = df.groupby("Prod")[["Week"]] \
          .apply(lambda x: x.value_counts("Week").add_prefix("Week").to_dict()).to_dict()

>>> dyear
{'A': {2018: 4}, 'B': {2018: 3}, 'C': {2018: 2}}

>>> dweek:
{'A': {'Week1': 4}, 'B': {'Week1': 3}, 'C': {'Week1': 2}}