pandas, reorder categorical

pandas. reorder cetegorical.py

# experiment with ordering and sorting
import pandas as pd
# create a categorical series
s = pd.Series(pd.Categorical(['a', 'b', 'c', 'a'], ordered=True))
print(s)
print(s.cat.categories)
print(s.cat.codes)

# set categories
print('-- set categories (changes the series values) --')
s2 = s.cat.set_categories(['c', 'a', 'b'])
print(s2)
print(s2.cat.categories)
print(s2.cat.codes)

# set categories (second method)
print('-- set categories 2 (does not change the series values) --')
s3 = s.copy(deep=True)
s3.cat.categories = ['c', 'a', 'b']
print(s3)
print(s3.cat.categories)
print(s3.cat.codes)

# set categories (third method)
print('-- set categories 3 (does not change the series values) --')
s4 = s.cat.set_categories(['c', 'a', 'b'], rename=True)
print(s4)
print(s4.cat.categories)
print(s4.cat.codes)

# reorder categories
print('-- reorder categories (changes the series values) --')
s5 = s.cat.reorder_categories(['c', 'a', 'b'])
print(s5)
print(s5.cat.categories)
print(s5.cat.codes)


# prints
# 0    a
# 1    b
# 2    c
# 3    a
# dtype: category
# Categories (3, object): ['a' < 'b' < 'c']
# Index(['a', 'b', 'c'], dtype='object')
# 0    0
# 1    1
# 2    2
# 3    0
# dtype: int8
# -- set categories (changes the series values) --
# 0    a
# 1    b
# 2    c
# 3    a
# dtype: category
# Categories (3, object): ['c' < 'a' < 'b']
# Index(['c', 'a', 'b'], dtype='object')
# 0    1
# 1    2
# 2    0
# 3    1
# dtype: int8
# -- set categories 2 (does not change the series values) --
# 0    c
# 1    a
# 2    b
# 3    c
# dtype: category
# Categories (3, object): ['c' < 'a' < 'b']
# Index(['c', 'a', 'b'], dtype='object')
# 0    0
# 1    1
# 2    2
# 3    0
# dtype: int8
# -- set categories 3 (does not change the series values) --
# 0    c
# 1    a
# 2    b
# 3    c
# dtype: category
# Categories (3, object): ['c' < 'a' < 'b']
# Index(['c', 'a', 'b'], dtype='object')
# 0    0
# 1    1
# 2    2
# 3    0
# dtype: int8
# -- reorder categories (changes the series values) --
# 0    a
# 1    b
# 2    c
# 3    a
# dtype: category
# Categories (3, object): ['c' < 'a' < 'b']
# Index(['c', 'a', 'b'], dtype='object')
# 0    1
# 1    2
# 2    0
# 3    1
# dtype: int8