python: filter pandas DataFrame getting all the rows in which in any of the values the string contai-CodePudding

I would like to get the number of rows in which any of the values contain a space

My starting point is:

li = [['erer','sfsddf', 'dfdf dfdf'],['erer','sfs/ df', 'dfdf dfdf'],['erer','sfsd/df', 'dfdf dfdf'],['er/er','sfsddf', 'dfdf dfdf'],['er/er','sfsddf', 'dfdf dfdf'],['er er','sfsddf', 'dfdf dfdf']]
df = pd.DataFrame(li, columns=[str(i) for i in range(3)])
df.apply(lambda col: col.str.contains(' ', na=False), axis=1)

(following: Filter pandas DataFrame by substring criteria)

But from that I dont know how to use that boolean DF to filter the original DF. What I want is to filter out all the rows where the boleans are False, i.e. not having space.

CodePudding user response：

For filtering rows contains space use DataFrame.any for boolean mask and filter in boolean indexing:

df1 = df[df.apply(lambda col: col.str.contains(' ', na=False), axis=1).any(axis=1)]
print (df1)
       0        1          2
0   erer   sfsddf  dfdf dfdf
1   erer  sfs/ df  dfdf dfdf
2   erer  sfsd/df  dfdf dfdf
3  er/er   sfsddf  dfdf dfdf
4  er/er   sfsddf  dfdf dfdf
5  er er   sfsddf  dfdf dfdf

If need count rows with space use Series.str.count with sum:

df['sum'] = df.apply(lambda col: col.str.count(' '), axis=1).sum(axis=1)
print (df)
       0        1          2  sum
0   erer   sfsddf  dfdf dfdf    1
1   erer  sfs/ df  dfdf dfdf    2
2   erer  sfsd/df  dfdf dfdf    1
3  er/er   sfsddf  dfdf dfdf    1
4  er/er   sfsddf  dfdf dfdf    1
5  er er   sfsddf  dfdf dfdf    2