14.6 Converting categorical data

The GENDER column contains categorical data rather than numeric. The items in the column belong to a fixed set of values, which are usually strings. In this case, the values are 'F' and 'M'. While we can check if an item is equal to one of these, it is often easier to convert the categorical column to multiple numeric “dummy columns” containing 0 and 1.

Here are the first two rows of df:

df.head(2)

          Locality  Postcode  Breed  Colour Gender
0  DANDENONG NORTH      3175  DOMSH     TAB      F
1  DANDENONG NORTH      3175  DOMLH  BLAWHI      M

and this is what we get when we use get_dummies on the GENDER column:

pd.get_dummies(df, columns=["Gender"]).head(2)

          Locality  Postcode  Breed  Colour  Gender_F  Gender_M
0  DANDENONG NORTH      3175  DOMSH     TAB         1         0
1  DANDENONG NORTH      3175  DOMLH ...