How to create sets for a column, separated by data types?

Question

I want to create a set in which only the string values of a column appear. The numerical values of the column are to be stored in another set.

Answer 1

You can use:

df['numeric'] = pd.to_numeric(df['col'], errors='coerce')
df['non-numeric'] = df['col'].mask(df['numeric'].notna())

Example:

   col  numeric non-numeric
0  abc      NaN         abc
1  123    123.0         NaN
2  123    123.0         NaN
3  NaN      NaN         NaN

Used input:

df = pd.DataFrame({'col': ['abc', 123, '123', float('nan')]})