python pandas - check if a string type exists in a column

Question

I know how to check for a 'nan' value in in column 'A' of dataframe 'df' as follows

df['A'].isnull().values.any()

but how can I check for a 'string', and I mean any string, since i do not know what the string text is, and then also to know which row it was found in?

Answer 1

If you are using Python 3, you can use a list comprehension and numpy.any

 import numpy as np

 np.any([isinstance(val, str) for val in df['A']])

If you are using Python 2, I believe that you need to replace str with basestring.

Answer 2

I would use vectorized Pandas approach:

Assuming we have the following DF:

In [116]: df = pd.DataFrame({'a':[1,2,'aaa', 3.14, 2.71], 'b':['2016-01-01', 'bbb', '2016-02-02', '2016-03-03', 'ZZZ']})

In [117]: df
Out[117]:
      a           b
0     1  2016-01-01
1     2         bbb
2   aaa  2016-02-02
3  3.14  2016-03-03
4  2.71         ZZZ

In [118]: df.dtypes
Out[118]:
a    object
b    object
dtype: object

check for strings in the column that supposed to be numeric :

In [119]: pd.to_numeric(df.a, errors='coerce')
Out[119]:
0    1.00
1    2.00
2     NaN
3    3.14
4    2.71
Name: a, dtype: float64

In [120]: pd.to_numeric(df.a, errors='coerce').isnull()
Out[120]:
0    False
1    False
2     True
3    False
4    False
Name: a, dtype: bool

In [121]: df.loc[pd.to_numeric(df.a, errors='coerce').isnull()]
Out[121]:
     a           b
2  aaa  2016-02-02

check for strings in the column that supposed to be datetime-like :

In [122]: pd.to_datetime(df.b, errors='coerce')
Out[122]:
0   2016-01-01
1          NaT
2   2016-02-02
3   2016-03-03
4          NaT
Name: b, dtype: datetime64[ns]

In [123]: df.loc[pd.to_datetime(df.b, errors='coerce').isnull()]
Out[123]:
      a    b
1     2  bbb
4  2.71  ZZZ