[FIXED] Replace transformation: What is the right way to do it?

Issue

I do this to replace a character in a string:

df['msg'] = df['msg'].str.replace(u'X','W')

And get this warning:

A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

Then, I try to do this same transformation the right way (I thought) to avoid that warning:

df.loc[:,'msg'] = df.loc[:,'msg'].str.replace(u'X','W')

But, I am still getting the same warning, even though both codes works fine.

What is the correct way to do this kind of transformation?

Solution

This warning can be resolved by using the method copy():

df.loc[:,'msg'] = df['msg'].str.replace(u'X','W').copy()

Or assign()

df = df.assign(msg=df['msg'].str.replace(u'X','W'))

Answered By - Baron Legendre