Relationship-Analysis-between-Packages

Research on the relationship between packages in Python.

以下为统计、分析结果：

Github仓库中packages使用情况统计

在通过爬虫获取的1000个仓库中，统计它们使用的packages，最常被使用的packages如下。其中，一些python自带的包（os, sys, time）最常被使用；第三方库中，numpy被使用频率最高；网页制作框架，数据可视化与数据处理相关的包，机器学习/深度学习相关的包等，出现频率也很高。

序号	package名	被使用次数
1	os	419
2	sys	366
3	time	183
4	numpy	180
5	re	136
6	json	118
7	setuptools	117
8	datetime	112
9	django	95
10	random	89
11	logging	89
12	matplotlib	83
13	requests	82
14	argparse	78
15	__future__	76
16	math	69
17	subprocess	64
18	pandas	63
19	unittest	59
20	urllib	57
21	collections	53
22	tensorflow	48
23	sklearn	47
24	flask	44
25	distutils	41
26	threading	39
27	scipy	37
28	shutil	36
29	csv	35
30	hashlib	33

同时出现在一个工程中的packages统计

在通过爬虫获取的1000个仓库中，统计它们使用的packages，最常被使用在同一个工程中的packages如下。其中，与系统相关的包（如，os,time,sys）经常出现；与科学计算，数据处理相关的包（如，numpy, matplotlib, pandas）也经常一起出现。

序号	package-1	package-2	同时出现次数
1	os	sys	283
2	os	time	118
3	os	re	101
4	time	sys	97
5	os	django	93
6	django	sys	91
7	numpy	matplotlib	86
8	re	sys	86
9	os	numpy	82
10	logging	os	73
11	os	datetime	70
12	os	json	70
13	os	setuptools	66
14	numpy	sys	66
15	datetime	sys	60
16	os	argparse	60
17	json	sys	57
18	os	__future__	57
19	time	re	55
20	os	subprocess	54
21	numpy	sklearn	53
22	time	datetime	52
23	time	json	51
24	pandas	numpy	50
25	logging	sys	49
26	os	random	49
27	sys	subprocess	48
28	argparse	sys	48
29	__future__	sys	48
30	sys	__future__	48
31	sklearn	matplotlib	46
32	os	requests	45
33	numpy	tensorflow	45
34	time	numpy	44
35	logging	time	43
36	os	matplotlib	42
37	math	sys	40
38	setuptools	sys	39
39	re	json	39
40	datetime	json	38

GitHub上仓库使用的许可证统计

在通过爬虫获取的16800个仓库中，使用许可证的统计情况（许可证项为None的在统计中会被忽略）如下。其中mit许可证占了超过三分之一的比例，被使用最多。

许可证名	仓库数	百分比
MIT License	4736	38.90%
Apache License 2.0	1558	12.80%
Other	1551	12.74%
BSD 3-clause "New" or "Revised" License	1478	12.14%
GNU General Public License v3.0	1157	9.50%
GNU General Public License v2.0	532	4.37%
BSD 2-clause "Simplified" License	433	3.56%
GNU Affero General Public License v3.0	191	1.57%
GNU Lesser General Public License v3.0	133	1.09%
The Unlicense	96	0.79%
ISC License	80	0.66%
GNU Lesser General Public License v2.1	67	0.55%
Do What The F*ck You Want To Public License	48	0.39%
Creative Commons Zero v1.0 Universal	40	0.33%
Mozilla Public License 2.0	37	0.30%
SIL Open Font License 1.1	12	0.10%
Eclipse Public License 1.0	6	0.05%
Creative Commons Attribution Share Alike 4.0	5	0.04%
zlib License	4	0.03%
Creative Commons Attribution 4.0	3	0.02%
Artistic License 2.0	2	0.02%
BSD 3-clause Clear License	2	0.02%
Open Software License 3.0	1	0.01%
European Union Public License 1.1	1	0.01%
PostgreSQL License	1	0.01%

GitHub上最常见的标签

在通过爬虫获取的16800个仓库中，最常被使用的标签如下。其中，深度学习相关标签（如，deep-learning, pytorch, tensorflow, neural-network, keras, gan等），占了很大一部分；网页制作框架（如，django, flask），也出现了较多次。

序号	标签	出现次数
1	python	2403
2	deep-learning	373
3	tensorflow	330
4	machine-learning	307
5	django	289
6	python3	188
7	pytorch	151
8	security	139
9	linux	128
10	flask	124
11	cli	99
12	keras	97
13	docker	93
14	nlp	90
15	data-science	82
16	natural-language-processing	77
17	computer-vision	76
18	python-3	72
19	asyncio	71
20	neural-network	69
21	api	68
22	visualization	64
23	reinforcement-learning	63
24	terminal	61
25	windows	60
26	javascript	56
27	bot	51
28	python-library	50
29	raspberry-pi	49
30	gan	49

标签中最常见的packages

序号	包名	出现次数
1	tensorflow	330
2	http	42
3	numpy	39
4	json	37
5	sqlalchemy	36
6	pandas	33
7	theano	32
8	jupyter	29
9	scikit-learn	28
10	html	23
11	ipython	21
12	email	20
13	opencv	20
14	matplotlib	20
15	parser	18
16	logging	18
17	requests	17
18	twisted	16
19	csv	16
20	aiohttp	16

最常组合出现在标签中的packages

序号	package-1	package-2	出现次数
1	theano	tensorflow	14
2	numpy	pandas	9
3	http	requests	8
4	jupyter	ipython	7
5	pandas	matplotlib	6
6	numpy	scipy	5
7	pyside	pyqt4	5
8	pandas	scikit-learn	5
9	scipy	numpy	5
10	numpy	opencv	4
11	numpy	tensorflow	4
12	json	html	4
13	html	json	4
14	opencv	tensorflow	4
15	pip	virtualenv	4
16	tensorflow	numpy	4
17	numpy	scikit-learn	3
18	csv	json	3
19	threading	multiprocessing	3
20	json	http	3

Stackoverflow36k个问题中包出现情况

序号	包名	出现次数
0	numpy	2229
1	pandas	1902
2	matplotlib	1584
3	string	797
4	scipy	654
5	regex	527
6	tkinter	478
7	pip	474
8	json	401
9	opencv	396
10	csv	386
11	sqlalchemy	378
12	datetime	359
13	subprocess	348
14	multiprocessing	345
15	tensorflow	340
16	virtualenv	297
17	scikit-learn	289
18	pyqt	276
19	ipython	264
20	html	261
21	nltk	223
22	logging	202
23	lxml	174
24	math	170
25	pyqt4	167
26	cython	161
27	pickle	149
28	pygame	142
29	random	139
30	argparse	133
31	ctypes	132
32	setuptools	127
33	wxpython	124
34	jinja2	115
35	twisted	115
36	time	107
37	email	104
38	tornado	95
39	ssl	89
40	distutils	84
41	py2exe	82
42	mysql-python	81
43	sqlite3	80
44	networkx	73
45	io	72
46	pymongo	69
47	pywin32	63
48	types	62
49	pyside	62
50	pygtk	57
51	kivy	57
52	queue	54
53	pyodbc	51
54	jupyter	51
55	itertools	49
56	cgi	48
57	pytz	41
58	nose	41
59	gevent	41
60	sympy	40
61	pdb	39
62	console	38
63	pyserial	38
64	bokeh	38
65	statsmodels	37
66	pillow	35
67	decimal	34
68	theano	34
69	copy	32
70	spyder	31
71	pytables	30
72	reportlab	28
73	numbers	27
74	rpy2	27
75	zipfile	26
76	pyyaml	25
77	base64	24
78	warnings	24
79	h5py	24
80	heatmap	23
81	pycurl	23
82	collections	22
83	cmd	22
84	pyparsing	22
85	gzip	21
86	struct	21
87	smtplib	20
88	keyword	19
89	simplejson	19
90	ftplib	19
91	configparser	19
92	readline	19
93	scikit-image	18
94	pyaudio	18
95	python-dateutil	18
96	numba	18
97	sys	17
98	select	17
99	gensim	17
100	mercurial	17

StackOverflow中同一问题下包出现情况：

序号	package1	package2	出现次数
1	numpy	scipy	790
2	numpy	pandas	430
3	matplotlib	numpy	384
4	matplotlib	pandas	288
5	matplotlib	scipy	146
6	pyqt4	pyqt	146
7	pip	virtualenv	138
8	csv	pandas	126
9	datetime	pandas	104
10	regex	string	92
11	cython	numpy	92
12	scikit-learn	numpy	90
13	opencv	numpy	86
14	setuptools	pip	64
15	matplotlib	ipython	62
16	pyqt	pyside	62
17	setuptools	distutils	60
18	scikit-learn	pandas	60
19	numpy	math	60
20	pandas	ipython	56
21	jupyter	ipython	54
22	csv	numpy	54
23	datetime	time	50
24	pandas	scipy	50
25	tkinter	matplotlib	44
26	pip	numpy	44
27	random	numpy	44
28	multiprocessing	numpy	42
29	matplotlib	heatmap	38
30	datetime	pytz	36
31	numpy	tensorflow	36
32	json	pandas	34
33	queue	multiprocessing	34
34	scikit-learn	scipy	34
35	statsmodels	pandas	30
36	pytables	pandas	30
37	pandas	string	30
38	numba	numpy	30
39	multiprocessing	pickle	28
40	networkx	matplotlib	28
41	scipy	math	28
42	simplejson	json	26
43	h5py	numpy	24
44	pyqt	matplotlib	24
45	datetime	numpy	24
46	sympy	numpy	24
47	email	smtplib	22
48	statsmodels	numpy	20
49	pip	distutils	20
50	pytables	numpy	20
51	pickle	numpy	20
52	pandas	sqlalchemy	20
53	distutils	cython	20
54	datetime	matplotlib	20
55	html	regex	20
56	ctypes	numpy	18
57	pyqt4	pyside	18
58	csv	json	16
59	opencv	matplotlib	16
60	lxml	html	16
61	ode	scipy	16
62	sympy	scipy	16
63	matplotlib	wxpython	16
64	setuptools	virtualenv	14
65	pip	lxml	14
66	numpy	ipython	14
67	html	pandas	14
68	pip	matplotlib	14
69	json	pickle	14
70	json	string	14
71	virtualenv	matplotlib	14
72	pyqt4	matplotlib	14
73	datetime	string	14
74	pymongo	json	14
75	html	jinja2	14
76	statsmodels	scipy	14
77	pip	mysql-python	12
78	multiprocessing	subprocess	12
79	py2exe	pyqt	12
80	tkinter	wxpython	12
81	spyder	matplotlib	12
82	random	string	12
83	nltk	tokenize	12
84	json	html	12
85	pip	tensorflow	12
86	opencv	scikit-image	12
87	regex	pandas	10
88	csv	string	10
89	pip	pygame	10
90	nltk	regex	10
91	types	pandas	10
92	pandas	heatmap	10
93	numpy	string	10
94	py2exe	wxpython	10
95	ctypes	pywin32	10
96	tkinter	pygame	10
97	sympy	matplotlib	10
98	io	pandas	10
99	scikit-learn	multiprocessing	10
100	pip	ipython	10

StackOverflow中热度最高问题排列：

                       

序号	包名	热度
0	string	69064758
1	pandas	45328071
2	numpy	43093597
3	matplotlib	36833422
4	pip	27232593
5	datetime	19288259
6	json	17297712
7	regex	12370782
8	subprocess	10981344
9	csv	10966890
10	scipy	10668052
11	tkinter	8859804
12	virtualenv	8447119
13	time	7351270
14	opencv	6600929
15	types	6246745
16	math	6126282
17	random	6106164
18	ipython	4965214
19	html	4889697
20	logging	4587029
21	sqlalchemy	4286895
22	multiprocessing	4051885
23	tensorflow	3794396
24	setuptools	3371261
25	scikit-learn	2989051
26	mysql-python	2891747
27	io	2805214
28	pyqt	2791496
29	nltk	2733686
30	copy	2600492
31	lxml	2521582
32	email	2394460
33	pickle	2291889
34	argparse	2197746
35	ssl	2039735
36	pygame	1924669
37	numbers	1911016
38	jinja2	1880599
39	keyword	1864734
40	pyqt4	1673453
41	decimal	1547834
42	wxpython	1525681
43	sqlite3	1339759
44	console	1313286
45	timeit	1238865
46	pyserial	1228376
47	ctypes	1100857
48	distutils	1071212
49	py2exe	1052789