CODEWITHSUNDEEP
pythondatetimepandas
pm1359
pm1359

Reputation: 632

pandas error when convert object to datetime

This is two lines of inputdataframe:

    ts  country       os product_id  total_users  total_purchases
0  0000-00-00   Brazil      iOS                       1                0
1  0000-00-00  Germany                                1                0

I have tried following commands in order to convert 'ts' which is object to datetime:

 df['ts'] = df['ts'].astype('datetime64[ns]')

and this is an error I've got: ValueError: Month out of range in datetime string "0000-00-00"

I know that there is a problem with 0000-00-00, but I don't know how to get ride of it and solve it to work?

Upvotes: 2

Views: 2002

Answers (1)

ℕʘʘḆḽḘ
ℕʘʘḆḽḘ

Reputation: 19375

easy peasy

df['ts'] = pd.to_datetime(df['ts'], errors ='coerce')

no need to clean the data. wrong timestamps will get the NaT (not a timestamp)

Upvotes: 4

Related Questions