Specifying conditional expressions with pandas dataframes

Question

I'd like to figure out how I can create a column in my dataframe based on a multiple check condition.

When I use a single conditional check this seems to work fine.

df['1/1/2017'] = np.where(df["Term 1 Start Date"] <= '1/1/2017'), 'True', 'False')

However when I introduce a second option to check upon this fails telling me that ValueError: The truth value of a Series is ambiguous.

df['1/1/2017'] = np.where(
    (df["Term 1 Start Date"] <= '1/1/2017' and df["Term 1 End Date"] > '1/1/2017'), 'True', 'False')

It should be noted that not all lines are filled in. The dates contain dates as you would expect. What can I do to have this new column populate based on two criteria?

current dataframe

Desired result

check the link https://stackoverflow.com/questions/21415661/logic-operator-for-boolean-indexing-in-pandas — BENY, Aug 24 '17 at 22:22

score 3 · Accepted Answer · answered Aug 24 '17 at 22:07

You'll need the bitwise operator & instead of and. You'll also need extra sets of parentheses to handle precedence.

df['1/1/2017'] = np.where(((df["Term 1 Start Date"] <= '1/1/2017')\ 
                           & (df["Term 1 End Date"] > '1/1/2017'),\
                     'True', 'False'))

Reason being, and doesn't play nice with dataframes, but the bitwise operators &, | and ~ are all overloaded for use in conditionals.

Specifying conditional expressions with pandas dataframes

1 Answers1