Pandas 2.0 pyarrow后端日期时间操作

Question

我使用 pyarrow 后端有以下 pandas dataframe 对象：

crsp_m.info(verbose = True)

out:
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 4921811 entries, 0 to 4921810
Data columns (total 87 columns):
 #   Column             Dtype               
---  ------             -----               
 0   permno             int64[pyarrow]      
 1   secinfostartdt     date32[day][pyarrow]
 2   secinfoenddt       date32[day][pyarrow]
 3   securitybegdt      date32[day][pyarrow]
 4   securityenddt      date32[day][pyarrow]

我想将这些天推迟到月底，类似于我使用 pandas 日期时间所做的事情：

crsp_m["date"] = pd.to_datetime(crsp_m.date)
crsp_m["date"] = crsp_m.date + pd.tseries.offsets.MonthEnd(0)

对象的等效操作是什么date32[day][pyarrow]？

Answer 1

Maybe just convert afterwards?

crsp_m["date"] = (pd.to_datetime(crsp_m.date) + pd.tseries.offsets.MonthEnd(0)).astype("date32[pyarrow]")