How to drop the columns in pandas with multiple condtions-CodePudding

I am new to python and pandas

On the below data frame ,I need to the drop the columns which are totally "None" , with "blanks and None", but not the columns with values and None

On the above table, I want Column A and C to be dropped because they are totally "None" or "blank and None", but Column B has some valid data at least in 3 cells, it should not be disturbed

how to give this condition in df.drop (pandas)

CodePudding user response：

You can test missing values NaN and None like Nonetype by DataFrame.isna, then possible strings by DataFrame.isin, chain by | for bitwise OR and pass to DataFrame.loc with invert mask for test if all values are Trues per columns (default axis=0) by DataFrame.all:

m = df.isna() | df.isin(['', 'None', 'none'])

df = df.loc[:, ~m.all()]

Or like comment, only in output are replaced values:

df = df.replace(['', 'None', 'none'],np.nan).dropna(axis=1, how='all')