删除熊猫中的“nan”行而不是“NaN”

Question

I need to delete the rows that contains 'nan%' in in the 'Precison' and 'Recall' columns, as below image shows,

I just need to remove all rows that shows 'nan%' both in 'Precision' and in 'Recall'. dropna() does not work here.

Answer 1

You can select all rows if not equal nan% in both columns:

df[df[['Precison','Recall']].ne('nan%').all(axis=1)]

Or you can replace all nan% to NaN for working DataFrame.dropna :

df = df.replace('nan%', np.nan).dropna(subset=['Precison','Recall'])