【數據分析可視化】數據分箱技術Binning

分箱：抽象理解爲蘋果根據大小不同分級分箱

import numpy as np
import pandas as pd
from pandas import Series,DataFrame

# 模擬成績分箱
score_list = np.random.randint(35, 100, size=20)
score_list

array([93, 35, 83, 44, 56, 62, 37, 86, 44, 82, 49, 91, 49, 82, 53, 89, 47,
       56, 38, 86])

# 成績評級分段
bins = [0,59,70,80,100]

# 分箱（返回Categories類型）
score_cut = pd.cut(score_list, bins)
score_cut

[(80, 100], (0, 59], (80, 100], (0, 59], (0, 59], ..., (80, 100], (0, 59], (0, 59], (0, 59], (80, 100]]
Length: 20
Categories (4, interval[int64]): [(0, 59] < (59, 70] < (70, 80] < (80, 100]]

# 每一個分箱多少人
pd.value_counts(score_cut)

(0, 59]      11
(80, 100]     8
(59, 70]      1
(70, 80]      0
dtype: int64

# 將模擬的成績 放入DataFrame
df = DataFrame()
df['score'] = score_list
df

	score
0	93
1	35
2	83
3	44
4	56
5	62
6	37
7	86
8	44
9	82
10	49
11	91
12	49
13	82
14	53
15	89
16	47
17	56
18	38
19	86

# 填充長度爲3的隨機字符串
df['student'] = [pd.util.testing.rands(3) for i in range(20)]
df

	score	student
0	93	8c1
1	35	cHy
2	83	6xy
3	44	6gY
4	56	tc5
5	62	r5T
6	37	3z3
7	86	vsy
8	44	F6h
9	82	hgC
10	49	xA9
11	91	iLZ
12	49	BVK
13	82	E9C
14	53	rbE
15	89	hSL
16	47	AIt
17	56	Gdk
18	38	AFX
19	86	JhU

# 利用pd.cut將數據處理並填充到DataFrame
pd.cut(df['score'], bins)

0     (80, 100]
1       (0, 59]
2     (80, 100]
3       (0, 59]
4       (0, 59]
5      (59, 70]
6       (0, 59]
7     (80, 100]
8       (0, 59]
9     (80, 100]
10      (0, 59]
11    (80, 100]
12      (0, 59]
13    (80, 100]
14      (0, 59]
15    (80, 100]
16      (0, 59]
17      (0, 59]
18      (0, 59]
19    (80, 100]
Name: score, dtype: category
Categories (4, interval[int64]): [(0, 59] < (59, 70] < (70, 80] < (80, 100]]

# cut的標籤化
df['Categories'] = pd.cut(df['score'], bins, labels=['low','ok','good','great'])
df

	score	student	Categories
0	93	8c1	great
1	35	cHy	low
2	83	6xy	great
3	44	6gY	low
4	56	tc5	low
5	62	r5T	ok
6	37	3z3	low
7	86	vsy	great
8	44	F6h	low
9	82	hgC	great
10	49	xA9	low
11	91	iLZ	great
12	49	BVK	low
13	82	E9C	great
14	53	rbE	low
15	89	hSL	great
16	47	AIt	low
17	56	Gdk	low
18	38	AFX	low
19	86	JhU	great