How to extract column labels only from pandas dataframe?

user_14690158 0
Viewed 863

I want to extract column labels only which has time format from data frame. My data looks like below

My data looks like this

and I want to get only data like 06-Dec-2020,07-Dec-2020,08-Dec-2020,09-Dec-2020,10-Dec-2020,11-Dec-2020,12-Dec-2020

Looking for great support

pandas dataframe python-3.x python
2 Answers

You can use filter with regex pattern:

df.filter(regex=r'\d{2}-\w{3,}-\d{4}').columns

Result:

Index(['06-Dec-2020', '07-Dec-2020', '08-Dec-2020', '09-Dec-2020',
       '10-Dec-2020', '11-Dec-2020', '12-Dec-2020'],
      dtype='object')
user_128331660

Assuming the dates might be across multiple months, years,

cols = ['blah', '06-Dec-2020', '20-Dec-2020','01-Jan-2021']
df = pd.DataFrame(columns = cols)

You can convert the dates to datetime with errors set to coerce which will convert non-dates to NaT. Then filter the columns using boolean indexing,

pd.to_datetime(df.columns, format = '%d-%b-%Y', errors='coerce').notna()

array([False,  True,  True,  True])

df.columns[pd.to_datetime(df.columns, format = '%d-%b-%Y', errors='coerce').notna()].tolist()

Output:

['06-Dec-2020', '20-Dec-2020', '01-Jan-2021']
user_62873080
Related