[FIXED] How to modify numercial values in a column of mixed data types in a pandas dataframe?

Issue

I have a pandas dataframe in python that looks like this (my actual dataframe is MUCH bigger than this):

  col_1 col_2
0   0.8   0.1
1  nope   0.6
2   0.4   0.7
3  nope  nope

How can I perform some operations on the numerical values of specific columns. For example, multiply the numerical values of col_2 by 10 to get something like this:

  col_1 col_2
0   0.8   1
1  nope   6
2   0.4   7
3  nope  nope

Although it looks like a simple task I couldn't find a solution for it anywhere on internet.

Thanks in advance.

Solution

First you need to convert the object type columns to numeric columns by using pd.to_numeric:

In [141]: df.col_2 = pd.to_numeric(df.col_2, errors='coerce')

errors='coerce' converts all non-numeric type values in the column to NaN.

Then, multiply by 10:

In [144]: df.col_2 = df.col_2 * 10

In [145]: df
Out[145]: 
  col_1  col_2
0   0.8    1.0
1  nope    6.0
2   0.4    7.0
3  nope    NaN

if you want to convert NaN back to nope, you can use df.fillna:

In [177]: df.fillna('nope', inplace=True)

In [178]: df
Out[178]: 
  col_1 col_2
0   0.8     1
1  nope     6
2   0.4     7
3  nope  nope

Answered By - Mayank Porwal