why date-time is automatically changed to the object datatype?

Question

Need to convert object type to Date-time and use that converted date-time is unknowingly changed to the object again while filtering.

msft['Tested On'] = pd.to_datetime(msft['Tested On'])
msft['Tested On'].dt.date
df = msft[msft['Tested On'] == '2018-02-02']

Actual result: TypeError: 'NoneType' object is not subscriptable

Expected result: Need to display the filtered list with the other rows and columns also

Do you assign `msft['Tested On'].dt.date` ? Can you add some data sample? — jezrael, Feb 08 '18 at 13:16
[link](https://stackoverflow.com/questions/48568016/how-to-filter-the-date-in-excel-using-pandas) Data in that link-page will help — gomathi subramanian, Feb 08 '18 at 13:21
I don't how add a table in stack overflow, that's why created a link. — gomathi subramanian, Feb 08 '18 at 13:22
@gomathisubramanian you can just attach `df_to_dict()` so people can reproduce your example quickly. — rpanai, Feb 08 '18 at 13:44

jezrael · Accepted Answer · 2018-02-08T13:38:35.037

0

I think you need to_datetime and if first digit is day, add parameter dayfirst=True:

print (msft)
  Title plan     run   status           Tested On
0     A   P0      P0   Passed  11/08/2018 8:42 AM
1     B   P0      P0   Failed  12/08/2018 8:42 AM
2     C    -  Canary   Passed  10/08/2018 8:42 AM
3     D    -  Sanity  Blocked  11/08/2018 8:42 AM

msft['Tested On'] = pd.to_datetime(msft['Tested On'], dayfirst=True)
print (msft)
  Title plan     run   status           Tested On
0     A   P0      P0   Passed 2018-08-11 08:42:00
1     B   P0      P0   Failed 2018-08-12 08:42:00
2     C    -  Canary   Passed 2018-08-10 08:42:00
3     D    -  Sanity  Blocked 2018-08-11 08:42:00

Then remove times by dt.floor and compare:

df = msft[msft['Tested On'].dt.floor('d') == '2018-08-11']
print (df)
  Title plan     run   status           Tested On
0     A   P0      P0   Passed 2018-08-11 08:42:00
3     D    -  Sanity  Blocked 2018-08-11 08:42:00

Detail:

print (msft['Tested On'].dt.floor('d'))
0   2018-08-11
1   2018-08-12
2   2018-08-10
3   2018-08-11
Name: Tested On, dtype: datetime64[ns]

edited Feb 08 '18 at 13:38

answered Feb 08 '18 at 13:24

jezrael

822,522
95
1,334
1,252

But when i i tried this with my code, i get an Empty DataFrame Columns: [Plan, Run, Status, Tested On] Index: [] – gomathi subramanian Feb 08 '18 at 13:34
With your real data? – jezrael Feb 08 '18 at 13:35
Ah now I get the Datetime type. but the result is getting the empty dataframe – gomathi subramanian Feb 08 '18 at 13:41
Can you check if data re converted to datetimes correctly? If not swapped day with months? – jezrael Feb 08 '18 at 13:42
`print (msft['Tested On'].dt.floor('d'))` - why d is for? – gomathi subramanian Feb 08 '18 at 13:42
it is day, it remove times – jezrael Feb 08 '18 at 13:42

why date-time is automatically changed to the object datatype?

1 Answers1