謨ｰ蟄励�蜃ｺ迴ｾ謨ｰ繧呈焚縺医ｋ
====================
繝励Ο繧ｰ繝ｩ繝�縺ｮ讎りｦ�
----------------
縺ゅｋ繧ｷ繧ｹ繝�Β縺檎函謌舌☆繧玖ｪ崎ｨｼ繧ｳ繝ｼ繝峨�6譯√�謨ｴ謨ｰ縺ｧ縺吶ゅ％縺ｮ繧ｳ繝ｼ繝峨↓迴ｾ繧後ｋ蜷�｡√�謨ｰ蟄�(Digit)縺ｮ蛻�ｸ�′縺ｩ縺ｮ遞句ｺｦ荳讒倥↑縺ｮ縺九ｒ隱ｿ縺ｹ縺ｾ縺吶�
螳滄圀縺ｫ逕滓�縺輔ｌ縺溘さ繝ｼ繝峨�荳隕ｧ縺九ｉ縲∝推譯√�謨ｰ蟄励↓蛻�ｧ｣縺励※鬆ｻ蠎ｦ繧呈焚縺医∪縺吶�
縺ｾ縺壹�逕滓�縺輔ｌ縺歡ode縺ｮ荳隕ｧ繧貞､画焚\ ``codes``\ 縺ｫ蜑ｲ繧雁ｽ薙※縺ｾ縺吶�
.. code:: ipython3
#!python3
#-*- coding:utf-8 -*-
# '#'莉･髯阪�陦梧忰縺ｾ縺ｧ縺後さ繝｡繝ｳ繝医↓縺ｪ繧翫∪縺吶ゑｼ医�繝ｭ繧ｰ繝ｩ繝�縺ｨ縺励※縺ｯ螳溯｡後＆繧後↑縺�ｼ�
nlen=6 # 繧ｳ繝ｼ繝峨�譯∵焚縺ｯ6譯�
codes=(
227524, 463251, 702567, 601620,
129458, 413239, 629380, 526093,
261547, 666552, 853626, 513298,
142167, 612906, 995697, 660500,
954404, 651454, 566439, 730975,
578967, 603424, 435636, 891667,
757294, 325567, 131075, 757309,
198547, 542718, 511525, 357679,
245280,
# 2021/05/14
40245, 314151, 689785, 246930, 429093,
546479, 240262, 631261, 559379, 869925,
814588, 766660, 588608, 815828, 946364,
542718, 710428, 725610, 863406, 250107,
629444, 865116, 86879,
# 012345 縺ｮ繧医≧縺ｫ0縺九ｉ蟋九∪繧議ode縺ｯ, 0繧堤怐縺�※譖ｸ縺上％縺ｨ縺ｫ縺励∪縺吶�
)
谺｡縺ｫ荳縺､縺ｮ code 縺ｮ蜷�｡√↓迴ｾ繧後ｋ謨ｰ蟄励�蜃ｺ迴ｾ蝗樊焚繧呈焚縺医∪縺吶�
荳逡ｪ荳九�譯√°繧蛾��分縺ｫ縺ｿ縺ｦ縺�″縺ｾ縺吶�
.. code:: ipython3
def count_digits(code,nlen=6): #縲髢｢謨ｰ縲`count_digits`繧貞ｮ夂ｾｩ縺励∪縺吶�
# 譛螟ｧ莠後▽縺ｮ蠑墓焚繧呈戟縺｡縺ｾ縺吶′縲∽ｺ後▽逶ｮ縺ｮ蠑墓焚`nlen`縺ｯ逵∫払蜿ｯ閭ｽ縺ｧ縲�
# 逵∫払縺輔ｌ縺溷�ｴ蜷医�蛟､(譌｢螳壼､)縺ｯ6縺ｧ縺吶�
acc={} #縺薙ｌ縺ｫ謨ｰ蟄玲ｯ弱�蜃ｺ迴ｾ蝗樊焚繧貞�繧後※縺�″縺ｾ縺吶�
code=code+10**nlen # 譛荳贋ｽ阪′0縺ｮ蝣ｴ蜷医ｒ蜿悶ｊ謇ｱ縺�◆繧√↓縲�10**nlen繧定ｶｳ縺励※縺翫￥縲�
# code =227524 縺ｧ縺ゅｌ縺ｰ縲code=1227524縺ｨ縺吶ｋ縺ｨ縺�≧縺薙→縲�
while code>=10:
code, d=divmod(code, 10) #蝠� `code` 縺ｨ菴吶ｊ `d` 縺ｫ蛻�ｧ｣縺励∪縺吶�
# code=1227524縲=> code=122752,d=4
acc[d]=acc.get(d,0)+1
# acc縺ｮkey縺ｫd縺後≠繧後�縺昴�蛟､縺ｫ1繧定ｶｳ縺励◆繧ゅ�縲√＆繧ゅ↑縺代ｌ縺ｰ 0+1繧誕cc[d]縺ｫ險ｭ螳壹☆繧九�
return acc
荳隕ｧ陦ｨ\ ``codes``\ 縺ｮ蜈ｨ縺ｦ縺ｮ\ ``code``\ 縺ｫ迴ｾ繧後ｋ謨ｰ蟄励↓鬆ｻ蠎ｦ繧呈焚縺医※縲∬ｶｳ縺励※縺�″縺ｾ縺吶�
邨先棡縺ｮ蠎ｦ謨ｰ蛻�ｸ�｡ｨ\ ``acc``\ 繧帝未謨ｰ縺ｮ蛟､縺ｨ縺励※霑斐＠縺ｾ縺吶�
.. code:: ipython3
def count_digits_in_list(codes,nlen):
acc={}
for code in codes: #codes縺ｮ荳ｭ縺ｫ蜈･縺｣縺ｦ縺�ｋ蜈ｨ縺ｦ縺ｮcode縺ｫ縺､縺�※
count=count_digits(code,nlen) #繧ｳ繝ｼ繝峨�荳ｭ縺ｮ謨ｰ蟄励�蜃ｺ迴ｾ陦ｨ繧剃ｽ懊ｋ縲�
for d in count: #code縺ｧ縺ｮ謨ｰ蟄励�蜃ｺ迴ｾ謨ｰ繧誕cc縺ｫ雜ｳ縺励※陦後￥縲�
acc[d]=acc.get(d,0)+count[d]
return acc
``codes``\ 縺ｮ荳ｭ縺ｮ謨ｰ蟄励�蠎ｦ謨ｰ蛻�｡ｨ繧貞魂蛻ｷ縺励√ヲ繧ｹ繝医げ繝ｩ繝�繧定｡ｨ遉ｺ縺輔○縺ｦ縺ｿ縺ｾ縺吶�
.. code:: ipython3
import matplotlib.pyplot as pyplot #繧ｰ繝ｩ繝穂ｽ懈�縺ｮ貅門ｙ縺ｮ縺翫∪縺倥↑縺�
import numpy
def main():
counts=count_digits_in_list(codes,nlen) #code縺ｮ荳隕ｧcodes縺九ｉ謨ｰ蟄励�蜃ｺ迴ｾ謨ｰ縺ｮ陦ｨ繧剃ｽ懈�縺吶ｋ縲�
print("digit: count")# 蜊ｰ蛻ｷ縺吶ｋ陦ｨ縺ｮ繝ｩ繝吶Ν繧貞魂蛻ｷ縺吶ｋ縲�
for d in sorted(counts): # 蜃ｺ迴ｾ陦ｨ縺ｫ縺ゅｉ繧上ｌ繧区焚蟄励ｒ謨ｴ蛻励＆縺帙◆鬆�分縺ｫ縲�
print (d,":",counts[d]) # 謨ｰ蟄�(d)縺ｨ縺昴�蜃ｺ迴ｾ謨ｰ(count[d])繧貞魂蛻ｷ縺吶ｋ縲�
# print("average:",len(codes)*nlen/10) #謨ｰ蟄励�蜃ｺ迴ｾ謨ｰ縺ｮ蟷ｳ蝮�､繧貞魂蛻ｷ縺吶ｋ縲�
print("average:", sum(counts.values())/len(counts)) # 6*len(codes)/10
print("std. dev:{:4.1f}".format(numpy.std(list(counts.values()))))
pyplot.bar(counts.keys(),counts.values()) # bar繧ｰ繝ｩ繝輔ｒ蜊ｰ蛻ｷ縺吶ｋ縲�
pyplot.draw() #繧ｰ繝ｩ繝穂ｽ懈�縺ｮ縺翫∪縺倥↑縺�
pyplot.show() #繧ｰ繝ｩ繝穂ｽ懈�縺ｮ譛蠕後�縺翫∪縺倥↑縺�. Jupyter Notebook縺ｧ縺ｯ荳崎ｦ�
.. code:: ipython3
#繝輔ぃ繧､繝ｫ繧偵さ繝槭Φ繝峨→縺励※螳溯｡後☆繧九◆繧√�縺翫∪縺倥↑縺�
if __name__ == "__main__":
main()
.. parsed-literal::
digit: count
0 : 28
1 : 30
2 : 37
3 : 23
4 : 36
5 : 45
6 : 50
7 : 30
8 : 27
9 : 30
average: 33.6
std. dev: 8.0
.. image:: CountDigits_files/CountDigits_8_1.png
pandas/DataFrame
----------------
Data Analysis蛻�㍽縺ｧ譛霑代ｈ縺丈ｽｿ繧上ｌ繧汽ataFrame繧ｪ繝悶ず繧ｧ繧ｯ繝医ｒ菴ｿ縺�→縲�
遏ｭ縺��繝ｭ繧ｰ繝ｩ繝�縺ｧ繝��繧ｿ縺ｮ繝ｪ繧ｹ繝医°繧牙ｺｦ謨ｰ蛻�｡ｨ繧貞叙蜃ｺ縺励√∪縺溘ヲ繧ｹ繝医げ繝ｩ繝�繧定｡ｨ遉ｺ縺輔○繧九％縺ｨ縺後〒縺阪∪縺吶�
.. code:: ipython3
import pandas # DataFrame縺御ｽｿ縺医ｋ繧医≧縺ｫpandas繧偵う繝ｳ繝昴�繝医＠縺ｦ鄂ｮ縺阪∪縺吶�
def split_digits(code,nlen=6):
acc=[] # 縺ゅｉ繧上ｌ縺滓焚蟄励ｒ鬆�分縺ｫ菫晏ｭ倥☆繧九�
code=code+10**nlen # 譛荳贋ｽ阪′0縺ｮ蝣ｴ蜷医ｒ蜿悶ｊ謇ｱ縺�◆繧�
while code>=10:
code,d=divmod(code,10)
acc.insert(0,d)
return acc
def split_codes():
acc=[]
for code in codes:
acc +=split_digits(code)
return acc
.. code:: ipython3
# codes縺ｮ繝��繧ｿ繧剃ｸ譯√�謨ｰ蟄励↓蛻�ｧ｣縺励√Μ繧ｹ繝医ｒ縺､縺上ｊ縺ｾ縺�
df=pandas.DataFrame(split_codes()).value_counts(sort=False)
print(df)
print("average:",numpy.average(df))
df.plot.bar()
.. parsed-literal::
0 28
1 30
2 37
3 23
4 36
5 45
6 50
7 30
8 27
9 30
dtype: int64
average: 33.6
.. parsed-literal::
<AxesSubplot:xlabel='0'>
.. image:: CountDigits_files/CountDigits_11_2.png
matplotlib.pyplot
-----------------
matplotlib.pyplot繧剃ｽｿ縺｣縺ｦ縲√ヲ繧ｹ繝医げ繝ｩ繝�縺ｮ陦ｨ遉ｺ縺ｨ蠎ｦ謨ｰ蛻�ｸ�ｒ豎ゅａ繧九％縺ｨ繧ょ庄閭ｽ縺ｧ縺吶�
縺ｾ縺壹ヲ繧ｹ繝医げ繝ｩ繝�繧剃ｽ懈�縺励∪縺吶�
.. code:: ipython3
counts, digits, chart = pyplot.hist(
split_codes(),
bins=range(0,11)
)
.. image:: CountDigits_files/CountDigits_13_0.png
matplotlib.hist()髢｢謨ｰ縺ｯ縲√ン繝ｳ豈弱�險域焚蛟､(count)縲｜in縺ｮ遶ｯ縺ｮ蛟､(digits)縺昴＠縺ｦ縲√ヲ繧ｹ繝医げ繝ｩ繝�縺ｮ繧ｪ繝悶ず繧ｧ繧ｯ繝医ｒ蛟､縺ｨ縺励※霑斐＠縺ｦ縺阪∪縺吶ゅ〒縺吶°繧峨∵ｬ｡縺ｮ繧医≧縺ｪ繧ｹ繧ｯ繝ｪ繝励ヨ縺ｧ縲∝推繝薙Φ縺ｮ繝ｩ繝吶Ν縺ｨ菫よ焚蛟､繧貞魂蛻ｷ縺励∪縺吶Ｅigits縺ｮ隕∫ｴ�謨ｰ縺ｯcounts縺ｮ隕∫ｴ�謨ｰ繧医ｊ1螟壹＞縺薙→縺ｫ豕ｨ諢上′蠢�ｦ√〒縺吶�
.. code:: ipython3
for p in (zip(digits,counts)):
print(p)
print("average:", sum(counts)/len(counts))
print(len(digits), ">",len(counts))
.. parsed-literal::
(0, 28.0)
(1, 30.0)
(2, 37.0)
(3, 23.0)
(4, 36.0)
(5, 45.0)
(6, 50.0)
(7, 30.0)
(8, 27.0)
(9, 30.0)
average: 33.6
11 > 10
谺｡縺ｫ縲∵焚蟄励�蜃ｺ迴ｾ蝗樊焚縺ｮ蛻�ｸ�ｒ繝励Ο繝�ヨ縺励※縺ｿ縺ｾ縺吶や�5窶昴♀繧医�窶�6窶昴�蜃ｺ迴ｾ蝗樊焚縺檎ｪ∝�縺励※縺�ｋ縺薙→縺�
隕九※蜿悶ｌ縺ｾ縺吶�
.. code:: ipython3
pyplot.hist(
counts,
histtype="stepfilled",
density=False,
align="left",
bins=range(34-3*8,34+3*8,2),
)
pyplot.gca().set_xlabel("count")
pyplot.gca().set_ylabel(u"occurence");
.. parsed-literal::
Text(0, 0.5, 'occurence')
.. image:: CountDigits_files/CountDigits_17_1.png