Map string values in Dataframe to integers using Python

Question

So I have a Dataset which has a column containing name of colors(red, blue, green) and I want to convert these string values into int/float to fit into classifier. I was planning on using Python dictionary with keys as name of color and values as numbers. This was my code:

color_dict = {'red':1, 'blue':2, 'green':3}
for i in train['column_name']:
     train['column_name'][i] = color_dict[i]
print(train['column_name'])

Sadly, this did not work. What should I do differently to make it work?

I assume you're using Pandas, so I added the pandas tag for you. If that's incorrect, you can edit to fix it. — wjandrea
– wjandrea, Commented Sep 20, 2021 at 4:34

onepan · Accepted Answer · 2021-09-19 20:35:31Z

2

The answer is in the question :)

train["column_name"] = train["column_name"].map(color_dict)

See the docs for map.

The reason your solution didn't work is a bit tricky. When you access a value like you did (using chained brackets), you're working on a copy of the DataFrame object. Instead, use train.loc[i, "column_name"] = color_dict[i] to set the a single value in a column. See here for more details.

answered Sep 19, 2021 at 20:35

onepan

9545 silver badges8 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Map string values in Dataframe to integers using Python

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related