Drop rows where all Boolean columns are False - Python

Question

How do I drop a row if all rows have the value False? (here it would mean dropping 'kiwi'.)

print(df.head())

>   concatenated   category    apple  banana  orange
> 0  apple_banana    two        True    True   False
> 1        apple     one        True   False   False
> 2        banana    one        False   True   False
> 3        kiwi      one        False   False  False

You haven't shown your definition of `df`. – Scott Hunter Jul 16 '20 at 10:34 — Scott Hunter, Jul 16 '20 at 10:34

badhusha muhammed · Accepted Answer · 2020-07-17T15:51:35.503

1

You can iterate through each row in the df and compare the column, if you get 3 False then you can drop the row,

 for index, row in df.iterrows():
    if not row["apple"] and  not row["banana"] and not row["orange"]:
        df.drop(index=index, inplace=True)

Try this also

  df1 = df[~(~(df['banana'])& ~(df['apple']) & ~(df['orange']))]

Answer for your last commend

  boolean_columns = ['apple', 'banana', 'orange']
  x = " & ".join('~('+x+')' for x in boolean_columns) 
  x = "~("+x+")"
  df1 = df.query(x)

Try this

edited Jul 17 '20 at 15:51

answered Jul 16 '20 at 11:57

badhusha muhammed

481
4
13

Is there a more RAM-efficient, general way? I have many columns and rows in reality. – arv Jul 16 '20 at 12:01
Could I do this using a list rather than specifying columns? – arv Jul 16 '20 at 12:28
You mean, column names as list..? – badhusha muhammed Jul 16 '20 at 12:30
Yep like boolean_columns = ['apple ', 'banana', 'orange'] – arv Jul 16 '20 at 12:32

Drop rows where all Boolean columns are False - Python

1 Answers1