I have two dataframes. I want to drop the values in first dataframe (default) after comparing with second dataframe (provided by user)

def_df = pd.DataFrame([['alpha','beta'],['gamma','delta']],index=['ab_plot',gd_plot])

               0    1
ab_plot     alpha   beta
gd_plot     gamma   delta
rk_plot     ray     kite


user_df = pd.DataFrame([10,20],index=['alpha','beta'])

        0
alpha   10
beta    20

I want to compare two dataframes and know the possible plots for given user data.

Expected answer

['ab_plot'] # since user has provided data for `'alpha','beta'`

My approach:

posble_plots_with_user_data = [True for x in posble_plots.values if x in df.columns]

Present answer:

TypeError: unhashable type: 'numpy.ndarray'

CodePudding user response：

If need test all values if match at least one value by index from user_df use DataFrame.isin with DataFrame.any and filter def_df.index:

#changed data
def_df = pd.DataFrame([['alpha','beta'],['gamma','beta']],index=['ab_plot','gd_plot'])

user_df = pd.DataFrame([10,20],index=['alpha','beta'])

posble_plots_with_user_data = def_df.index[def_df.isin(user_df.index).any(axis=1)].tolist()
print (posble_plots_with_user_data)
['ab_plot', 'gd_plot']

If need rows with match all values per rows use DataFrame.all:

posble_plots_with_user_data = def_df.index[def_df.isin(user_df.index).all(axis=1)].tolist()
print (posble_plots_with_user_data)
['ab_plot']

Details:

print (def_df.isin(user_df.index))
             0     1
ab_plot   True  True
gd_plot  False  True