I have a dataframe
colA ColB
0 2020-11-30 00:00:00 546
1 Vendor account statement yrt22
2 :doflat: ttys7
3 2020-1-30 00:00:00 970
I want to filter 'colA' dates only (any date) other then dates not required. Required dataframe is as below.
colA ColB
0 2020-11-30 00:00:00 546
1 2020-1-30 00:00:00 970
If need datetimes in output replace non datetimes no misisng values and rmeove rows:
df['colA'] = pd.to_datetime(df['colA'], errors='coerce')
df = df.dropna(subset=['colA'])
If need original values test for values possible converted, so here not missing values:
df = df[pd.to_datetime(df['colA'], errors='coerce').notna()]
You may use:
df['colA'].str.contains('^\d+-\d+-\d+ \d{2}:\d{2}:\d{2}$')
The regex pattern ^\d+-\d+-\d+ \d{2}:\d{2}:\d{2}$ should match the datetime/timestamp values shown which you want to retain, see the demo:
try this:
df[[type(i)==pd.Timestamp for i in df['colA']]]
