CODEWITHSUNDEEP
pythonpandas
nomnom3214
nomnom3214

Reputation: 259

how to categorize a column and put in a new column in pandas

I have this dataframe where the type columns shows the type of a stuff

id    type     total
1     shoes      2
1     sandal     1
1     vest       2
1     tshirt     2
1     345        3
1     345        2

based on the type column, I want to categorize the type and create a new column called category. The rules are as follow:

- shoes and sandal as shoes
- vest, we keep it as vest
- tshirt, we also keep it as tshirt
- other than that, I want to keep it as other

so the desired result will be as follow:

id    type     total    category
1     shoes      2       shoes
1     sandal     1       shoes
1     vest       2       vest
1     tshirt     2       tshirt
1     345        3       other
1     345        2       other

how can I make it with python?

thanks in advance

Upvotes: 1

Views: 774

Answers (1)

U13-Forward
U13-Forward

Reputation: 71560

Try using map on a dictionary with fillna:

>>> df['category'] = df['type'].map({'sandal': 'shoes', 'shoes': 'shoes', 'vest': 'vest', 'tshirt': 'tshirt'}).fillna('other')
>>> df
   id    type  total category
0   1   shoes      2    shoes
1   1  sandal      1    shoes
2   1    vest      2     vest
3   1  tshirt      2   tshirt
4   1     345      3    other
5   1     345      2    other
>>>

As mentioned in the documentation of both pages (links shown above), map maps and replaces the values in the Series that are the keys of the dictionary to the values of the dictionary. Whereas if there are values in there series that aren't mapped in the dictionary, it would give NaNs, so you could just do fillna to replace the NaNs with others.

Upvotes: 6

Related Questions