Pandas column dtype is object but python thinks it is float

Question

I read in a csv like this

df = pd.read_csv(self.file_path, dtype=str)

then I try this:

 df = df[df["MY_COLUMN"].apply(lambda x: x.isnumeric())]

I get an AttributeError:

AttributeError: 'float' object has no attribute 'isnumeric'

Why is this happening? The column contains mostly digits.

I want to filter out the ones where there are no digits.

This question is not how to achieve that or do it better but why do I get an AttributeError here?

Answer 1

Why is this happening?

I think because NaN is not converting to string if use dtype=str , still is missing value, so type=float

Use Series.str.isnumeric for working isnumeric with missing values like all text functions in pandas:

df[df["MY_COLUMN"].str.isnumeric()]