CODEWITHSUNDEEP
pythonpandas
Fluxy
Fluxy

Reputation: 2978

How to map values and assign many values to "other" category?

I have the following DataFrame df:

col1  col2
3     A
3     B
5     A
1     C
6     B

How can I replace 3 by 0 and 5 by 1, while the rest of numbers should be replaced by 2?

Expected result:

col3  col2
0     A
0     B
1     A
2     C
2     B

This is what I wrote so far:

vals = {3:0, 5:1}
df["col3"] = df["col1"].map(vals)

Upvotes: 0

Views: 32

Answers (2)

Karn Kumar
Karn Kumar

Reputation: 8816

You already close to what you need, just use df.assign to create a new column and replace nan values with DataFrame.replace and convert later to int.

>>> df.assign(col3=df["col1"].map(vals).replace(np.nan, 2).astype(int))
   col1 col2  col3
0     3    A     0
1     3    B     0
2     5    A     1
3     1    C     2
4     6    B     2

On the top of it as you are looking to have only col2 and col3 hence you can drop col1 as follows ..

Expected output:

>>> df.assign(col3=df["col1"].map(vals).replace(np.nan, 2).astype(int)).drop(columns={'col1'})
  # df = df.assign(col3=df["col1"].map(vals).replace(np.nan, 2).astype(int)).drop(columns={'col1'})
  col2  col3
0    A     0
1    B     0
2    A     1
3    C     2
4    B     2

Upvotes: 1

jezrael
jezrael

Reputation: 862641

Because map return missing values for non matched values, add Series.fillna:

vals = {3:0, 5:1}
df["col3"] = df["col1"].map(vals).fillna(2).astype(int)
print (df)
   col1 col2  col3
0     3    A     0
1     3    B     0
2     5    A     1
3     1    C     2
4     6    B     2

Upvotes: 2

Related Questions