Compare two date columns and validate if DateA is earlier than DateB

Question

I have two columns with dates(mm/dd/yy). I need to validate if DateColumn_A comes before DateColumn_B.

I used the following code and I got this error "TypeError: must be string, not Series". Please help a beginner.

Code:

Column_A = datetime.strptime(df['DateColumn_A'], '%m %d %y')
Column_B = datetime.strptime(df['DateColumn_B'], '%m %d %y')

for index, row in dataframe.iterrows():
    if row[Column_A] < row[Column_B]
        print (index,row[Column_A])
    else:
        pass

Answer 1

You can compare them like this:

from datetime import datetime

Column_A = datetime.strptime(df['DateColumn_A'], '%m %d %y').date()
Column_B = datetime.strptime(df['DateColumn_B'], '%m %d %y').date()
diff = Column_A - Column_B
if diff > 0:
    # Column_A is greater than Column_B
else:
    # Column_B is greater than Column_A

Answer 2

Elaborating on my comment above with an example.

First make sure the date columns you are comparing are actually dates. You can do that using the pandas to_datetime function like so:

>>> df = df.apply(pd.to_datetime, errors='ignore')
>>> df.DateColumnA
0   2018-01-01
1   2018-05-01
Name: DateColumnA, dtype: datetime64[ns]

The below snippet is using boolean indexing. So df['DateColumnA'] < df['DateColumnB'] returns a series of Trues and Falses. And then df.loc[df['DateColumnA'] < df['DateColumnB']] is akin to saying "Give me the subset of the DataFrame where this condition is True"

 >>> df
  DateColumnA DateColumnB
0  2018-01-01  2018-02-01
1  2018-05-01  2018-01-01
>>> df['DateColumnA'] < df['DateColumnB']
0     True
1    False
dtype: bool
>>> df.loc[df['DateColumnA'] < df['DateColumnB']]
  DateColumnA DateColumnB
0  2018-01-01  2018-02-01