Convert Columns to categoricals¶

In [1]:

import pandas as pd
import janitor
from numpy import nan

In [2]:

df = pd.DataFrame({'col1': [2.0, 1.0, 3.0, 1.0, nan],
                   'col2': ['a', 'b', 'c', 'd', 'a'],
                   'col3': ['2020-01-01',
                   '2020-01-02',
                   '2020-01-03',
                   '2020-01-04',
                   '2020-01-05']})

df

Out[2]:

	col1	col2	col3
0	2.0	a	2020-01-01
1	1.0	b	2020-01-02
2	3.0	c	2020-01-03
3	1.0	d	2020-01-04
4	NaN	a	2020-01-05

In [3]:

df.dtypes

Out[3]:

col1    float64
col2     object
col3     object
dtype: object

Specific columns can be converted to category type:

In [4]:

cat = df.encode_categorical(column_names=['col1', 'col2', 'col3'])

In [5]:

cat.dtypes

Out[5]:

col1    category
col2    category
col3    category
dtype: object

Note that for the code above, the categories were inferred from the columns, and is unordered:

In [6]:

cat['col3']

Out[6]:

0    2020-01-01
1    2020-01-02
2    2020-01-03
3    2020-01-04
4    2020-01-05
Name: col3, dtype: category
Categories (5, object): ['2020-01-01', '2020-01-02', '2020-01-03', '2020-01-04', '2020-01-05']

Explicit categories can be provided, and ordered via the `kwargs`` parameter:

In [7]:

cat = df.encode_categorical(
            col1 = ([3, 2, 1, 4], "appearance"),
            col2 = (['a','d','c','b'], "sort")
            )

In [8]:

cat['col1']

Out[8]:

0      2
1      1
2      3
3      1
4    NaN
Name: col1, dtype: category
Categories (4, int64): [3 < 2 < 1 < 4]

In [9]:

cat['col2']

Out[9]:

0    a
1    b
2    c
3    d
4    a
Name: col2, dtype: category
Categories (4, object): ['a' < 'd' < 'c' < 'b']

When the order parameter is appearance, the categories argument is used as-is; if the order is sort, the categories argument is sorted in ascending order; if order is None``, then thecategories` argument is applied unordered.

A User Warning will be generated if some or all of the unique values in the column are not present in the provided categories argument.

In [10]:

cat = df.encode_categorical(col1 = ([4, 5, 6], "appearance"))

/workspaces/pyjanitor/janitor/functions/encode_categorical.py:131: UserWarning: 
                     None of the values in col1 are in
                     [4, 5, 6];
                     this might create nulls for all values
                     in the new categorical column.
                     
  categories_dict = _as_categorical_checks(df, **kwargs)

In [11]:

cat['col1']

Out[11]:

0    NaN
1    NaN
2    NaN
3    NaN
4    NaN
Name: col1, dtype: category
Categories (3, int64): [4 < 5 < 6]