Pandas date and text condition

Question

Given a df:

Date1                   Text1    
2018-03-20 00:00:00     abc       
2018-04-01 00:00:00     abc
2018-01-01 00:00:00     abc
2018-04-01 00:00:00     xyz
                        abc

My goal to add a new column where: if text = "abc" and Date1 is with 90 days from now then "New" The output would be:

Date1                   Text1    NewText
2018-03-20 00:00:00     abc      New
2018-04-01 00:00:00     abc      New
2018-01-01 00:00:00     abc
2018-04-01 00:00:00     xyz
                        abc

This is what I have:

days90 = date.today() - timedelta(90)

df['NewText'] = np.where(df['Text1'] = "abc" & df['Date1'] < pd.to_datetime(days90), "New", np.nan)

However, I'm keep running into errors.
AttributeError: Can only use .str accessor with string values, which use np.object_ dtype in pandas

Any suggestion? Many thanks!

Suggestion to try: 'df['NewText'] = np.where(np.logical_and(df['Text1'] == "abc", df['Date1'] < pd.to_datetime(days90)), "New", np.nan)' — gxpr, Apr 10 '18 at 21:33
df['NewText'] = np.where((df['Text1'] =="abc" )& ((pd.to_datetime('today') -df['Date1'] ).dt.days< 90), "New", np.nan) — BENY, Apr 10 '18 at 21:33

score 1 · Accepted Answer · answered Apr 11 '18 at 00:12

There are 3 errors in your code:

Test equality of variables via == operator.
Make sure you bracket individual conditions to avoid chained comparisons.
For a date to be within 90 days, check your test data is greater than days90.

Combined, the following code will work:

df['NewText'] = np.where((df['Text1'] == "abc") & (df['Date1'] > days90), "New", np.nan)

Pandas date and text condition

1 Answers1