Python: Pandas Dataframe How to Convert timedelta column to float column

Question

I have a column in a dataframe that has datatype [timedelta64[ns]]. I'm trying to convert it to a float.

This is a sample table:

ColA
227 days 00:00:00.000000000
316 days 00:00:00.000000000
226 days 00:00:00.000000000
153 days 00:00:00.000000000

Below is my desired table with datatype as float:

ColA
227 
316  
226  
153  

This is the code I tried:

df_EVENT5_24['ColA'] = df_EVENT5_24['ColA'].astype(float)

This is the error: TypeError: cannot astype a timedelta from [timedelta64[ns]] to [float64]

Answer 1

您可以使用apply和lambda来访问属性days （ https://pandas.pydata.org/pandas-docs/stable/timedeltas.html ）：

df_EVENT5_24['ColA'] = df_EVENT5_24.apply(lambda row: row.ColA.days, axis=1)

Answer 2

We can use the dt.days attribute. Example ColA is the name of your timedelta Series.

ColA = ColA.dt.days

To convert into a float, you need either df.astype

ColA = ColA.astype(float)

***ColA = ColA.dt.days.astype(float)***