| CARVIEW |
?mNumPy??????n???Ђ??}?ƃq?X?g?O???????g???ăf?[?^?Z?b?g?????????Ă݂悤?FPython?f?[?^????????
NumPy??Matplotlib??g?ݍ??킹?A?f?[?^?Z?b?g?Ɋ܂܂?Ă???f?[?^???ǂ̂悤?ȕ??z?ɂȂ??Ă??邩?????????Ă݂܂??傤?B?V???Ȓm?????????邩??????܂???B
???̋L???͉??????ł??B????o?^?i?????j????ƑS?Ă????????????܂??B
?A?ڊT?v
?@?{?A?ڂ?Python?ɂ??Ă̒m???????ɂ?????x?͐g?ɕt???Ă??????ΏۂƂ??āAPython?Ńf?[?^???????s????ŕK?{?Ƃ???????NumPy??pandas?AMatplotlib?Ȃǂ̊e?탉?C?u?????̊?{?I?Ȏg???????w??ł??????̂ł??B?????āA?????̎g????????????x?o??????ŁA?????????p???ăf?[?^???????s?????߂̑????ݏo?????Ƃ?ړI?Ƃ??Ă??܂??B
?@?O????California Housing?f?[?^?Z?b?g?i?J???t?H???j?A?̏Z??i?̃f?[?^?Z?b?g?j???ނƂ??āA?ő?l??ŏ??l?A???ϒl?A?????l?Ȃǂ̊?{???v?ʂ??v?Z???A2???s????f?[?^?Z?b?g???ǂ̂悤?ȓ??????????Ă???̂??????Ă݂܂????B????͓????f?[?^?Z?b?g???O???t?Ƃ??ĉ??????āA????Ȃ??????T???Ă݂邱?Ƃɂ??܂??傤?B
CSV?t?@?C???̓ǂݍ???
?@?O???scikit-learn?Ɋ܂܂?Ă???f?[?^?Z?b?g??ǂݍ???ŁA?????ϐ??i?????ʁj?ƖړI?ϐ??i???????x???j??2?????z??ɂ܂Ƃ߂??Ƃ???ŁACSV?t?@?C???Ƃ??ĕۑ????Ă??܂????B?????ŁA?????ł͂??̃f?[?^??ǂݍ??ނ??Ƃɂ??܂??BNumPy?ł?CSV?t?@?C???̓ǂݍ??݂ɂ?numpy.loadtxt?????g???܂??B
?@numpy.loadtxt???̍\?????ȉ??ɂ܂Ƃ߂܂??i?????j?B?ڍׂɂ??Ă?NumPy?̃h?L???????g?unumpy.loadtxt?v???Q?Ƃ??Ă????????B
numpy.loadtxt(fname, dtype=<class 'float'>, comments='#',
delimiter=None, skiprows=0, usecols=None)
?@?p?????[?^?[?i?????j?ɂ͈ȉ????w?肵?܂??B
- fname?F?ǂݍ??ރt?@?C???̖??O?B?w?肪?K?{
- dtype?F?t?@?C??????ǂݍ??f?[?^????ɍ쐬???鑽?????z??̃f?[?^?^?B?f?t?H???g?l??float?^
- comments?F?R?????g?̎n?܂??\???????̎w??B?f?t?H???g?l??'#'?B#?Ŏn?܂?s?̓R?????g?s?ł???A???̍s?̓??e?͌??ʂ̑??????z??ɂ͊܂܂?Ȃ??B?e?s?̍Ō?ɃR?????g??t?????邱?Ƃ??\
- delimiter?F?ǂݍ??ރt?@?C???ɂ?????f?[?^?̋?蕶???B?f?t?H???g?l??None?ł???A????͋???????蕶???ł??邱?Ƃ??Ӗ?????
- skiprows?F?ǂݍ??ރt?@?C???̐擪???牽?s??ǂݔ???????w??B?f?t?H???g?l??0?i?ǂݔ???Ȃ??j?B??q?̒ʂ?A?R?????g?s???ǂݔ?????
- usecols?F?ǂݍ??ށi???ʂ̑??????z??Ɋ܂߂?j??̎w??B?f?t?H???g?l??None?ł???A????͑S?Ă̗?̃f?[?^??ǂݍ??ނ??Ƃ??Ӗ?????
?@?O??ɏЉ??numpy.savetxt???ɂ??Ă????l?ɂ܂Ƃ߂Ă????܂??B????ɂ??Ă??ڍׂ?NumPy?̃h?L???????g?unumpy.savetxt?v???Q?Ƃ??Ă????????B
numpy.savetxt(fname, X, delimiter=' ', header='')
?@savetxt???̃p?????[?^?[?i?????j?ɂ͈ȉ????w?肵?܂??B
- fname?F???????ݐ?̃t?@?C???̖??O?B?w?肪?K?{
- X?F???????ރf?[?^?i1?????z??܂???2?????z??j?B?w?肪?K?{
- delimiter?F?f?[?^?̋?蕶???̎w??B?f?t?H???g?l?͋????i' '?j
- header?F?t?@?C???擪?ɕt??????w?b?_?s?̓??e?B?f?t?H???g?l?͋???i''?j
?@?O??́utest.csv?v?Ƃ??????O?̃t?@?C???Ɏ??̂悤?Ƀf?[?^??ۑ????Ă??܂????B
np.savetxt('test.csv', data, header=','.join(columns), delimiter=',')
?@?w?b?_?s?̓??e??'MedInc'?A'HouseAge'?Ȃǂ̐????ϐ??ƖړI?ϐ??ł???'MedHouseVal'???J???}?u,?v?ŋ???????̂ł??B???????A?擪??'#'???t??????A?S?̂??R?????g?s?ƂȂ??Ă??܂??B
?@?O??ɕۑ??????f?[?^??ǂݍ??ނɂ́A?ȉ??̃R?[?h?????s???܂??B
import numpy as np
data = np.loadtxt('test.csv', delimiter=',')
?@???邢?͐擪?s?̓ǂݍ??݂??X?L?b?v????ȉ??̃R?[?h?ł??\???܂???i???A?擪?s?͏?ŏq?ׂ??悤?ɃR?????g?s?ɂȂ??Ă???̂ŁAskiprows?p?????[?^?[?ɒl???w?肵?Ȃ??Ƃ??ǂ݂̂??ǂݔ????܂??j?B
data = np.loadtxt('test.csv', delimiter=',', skiprows=1)
?@??????ɂ???Anumpy.loadtxt???ɂ??A?O??ɕۑ??????f?[?^?̕??????ł??܂????B???ɑO??Ɠ??l??'MedInc'?Ȃǂ̕????????ēǂݍ??f?[?^?̗?ɃA?N?Z?X?ł???悤?ɁA?ǂݔ?????w?b?_?s??ǂݍ??݁A???????쐬???Ă????܂??傤?B
with open('test.csv') as f:
header = f.readline()
columns = header.replace('# ', '').strip().split(',')
col = {k: v for v, k in enumerate(columns)}
print(col)
# ?o?͌??ʁF
# {'MedInc': 0, 'HouseAge': 1, 'AveRooms': 2, 'AveBedrms': 3, 'Population': 4,
# 'AveOccup': 5, 'Latitude': 6, 'Longitude': 7, 'MedHouseVal': 8}
?@?ŏ???with???ł?test.csv?t?@?C??????w?b?_?s??1?s?????ǂݍ???ł??܂??B???̌?̃R?[?h?ł͓ǂݍ??w?b?_?s???玫?????쐬???Ă??܂??B
?l???ʐ??????߂Ă݂?
?@?????āA?悤?₭?O??̍Ō?Ɍ??Ă??????????R?[?h?̓o??ł??B
for k, v in col.items():
min_v = np.min(data[:, v])
percentile = np.percentile(data[:, v], [25, 50, 75])
max_v = np.max(data[:, v])
print(f'{k} quantile: {min_v, str(percentile), max_v}')
?@???̃R?[?h?????s????ƁA?ȉ??̂悤?ȏo?͂??????܂??B
?@???̃R?[?h?͊e??̍ŏ??l?ƍő?l?A???ꂩ??l???ʐ??ƌĂ??l??\????????̂ł??B?l???ʐ??Ƃ́A?f?[?^?̐??ɒ??ڂ??āA????4????1?A2????1?A4????3?ɓ????鋫?E???ǂ??ɂ??邩?????߂????̂ł??B????̓f?[?^?̕??z?Ƃ???ׂ?̂ɖ𗧂??܂??B??̃R?[?h?ł͗ƂɁAnumpy.percentile?????Ăяo???Ă??܂????A????̓p?[?Z???^?C???ƌĂ??l?????߂???̂ŁA?????25?i???j?A50?i???j?A75?i???j???^?v???ɂ܂Ƃ߂????̂?n?????ƂŎl???ʐ??????߂Ă??܂??B
?@?????āA?ŏ???MedInc?i?????̒????l?j?̌??ʂɂ??Ă????A?f?[?^??0.4999????15.0001?͈̔͂ɕ??z???Ă??āA???ʂ?4????1?ɓ?????f?[?^??0.4999????2.5634?ɁA????4????1?ɓ?????f?[?^??2.5634????3.5348?ɁA???̎???4????1?ɓ?????f?[?^??3.5348????4.74325?ɁA?ŏ?ʂ?4????1?ɓ?????f?[?^??4.74325????15.0001?ɕ??z???Ă???Ƃ????킯?ł??B?ʼn??ʂƍŏ?ʂ??????????̃f?[?^?????݂???̂?2.5634????4.74325?̊ԂƋɂ߂ċ????͈͂ł????A?ő?l??15.0001?Ƃ??Ȃ?傫?Ȓl?ƂȂ??Ă??邱?Ƃ??珊?????????Ȃ?ɂ?āA???̏????̕????L???Ȃ??Ă???Ƃ????????ł??B
?@???̂悤?ȃf?[?^?̕??z?̎d???́A?????ŕ\??????????O???t?ɂ?????????????₷?????ł??ˁB?????ō????Matplotlib?ƌĂ?郉?C?u???????g???āA?????????????Ă݂邱?Ƃɂ??܂??傤?BMatplotlib?ɂ??Ă͖{?V???[?Y?Ō???A??????Ƃ???????A?ڂ??s???\??Ȃ̂ŏڂ????g?????͏ȗ????܂????A?????Python?Ńf?[?^?????????邽?߂̃??C?u?????ł??BPython?????n?ɕW???ŕt????????̂ł͂Ȃ??̂ŁA?g?p????ɂ͎??O?Ɂupip install matplotlib?v?upy -m pip install matplotlib?v?Ȃǂ̃R?}???h?????s??????ŁA?ȉ??̃R?[?h?ŃC???|?[?g???Ă????K?v??????܂??B
import matplotlib.pyplot as plt
?@matplotlib.pyplot???W???[???ɂ͊ȒP?Ƀf?[?^?????????邽?߂̋@?\???܂܂?Ă??܂??B????͂??̒????????̊????g???Ă݂邱?Ƃɂ??܂??傤?B
Matplotlib?ɂ?锠?Ђ??}?̕`??
?@?ŏ??l?A?l???ʐ??A?ő?l?̑g??????????ɂ́u???Ђ??}?v?ƌĂ??O???t???g???܂??BMatplotlib?ł?matplotlib.pyplot.boxplot?????Ăяo???????ŁA?ȒP?ɂ?????s???܂??B?ȉ??̃R?[?h?????Ă????????B
plt.boxplot(data[:, col['MedInc']])
plt.show()
?@??قǂ͎????Ŏl???ʐ??????߂Ă??܂??????Aboxplot???ɂ͔??Ђ??}???쐬???????f?[?^?i?????ł?MedInc??j??n???????ł??B?ȉ??????̎??s???ʂł??B
?쐬???ꂽ???Ђ??}?@???Ђ??}?͍ŏ??l?ƍő?l?ɉ?????`???A??1?l???ʐ??Ƒ?3?l???ʐ??͈̔͂ň͂?ŁA?????l?͂??̔??̒??ɉ????ŕ`?????Ƃ??悭????܂??i??????㉺?ɐL?т????????u?Ђ??v?ɂȂ?܂??j?B?Ⴆ?AHouseAge????ɔ??Ђ??}??`???Ǝ??̂悤?ɂȂ?܂??B
HouseAge??̔??Ђ??}?@HouseAge??ł͊e?l?͎??̂悤?ɂȂ??Ă??܂????B
- ?ŏ??l?F1.0
- ??1?l???ʐ??F18
- ?????l?F29
- ??3?l???ʐ??F37
- ?ő?l?F52
?@?}?Ɣ?ׂ?Ƃ??????????̕ӂ?ɉ????┠?A???????Ȃ??Ђ????`????Ă??邱?Ƃ???????܂??B???????A??قǂ?MedInc??̔??Ђ??}?͂???Ƃ͂?????ƈقȂ??Ă??܂??BMedInc??ɂ??Ă͊e?l?͎??̒ʂ?ł????B
- ?ŏ??l?F0.4999
- ??1?l???ʐ??F2.5634
- ?????l?F3.5348
- ??3?l???ʐ??F4.74325
- ?ő?l?F15.0001
?@???̂??Ƃ???l????ƍő?l?̉????͐}?̈?ԏ?̕ӂ?ɂ??肻???Ȃ??̂ł????A???ۂɂ?8?????????????ӂ?ɉ???????????A???̏?ɂ͉????f?[?^?????????݂??Ă???悤?Ɍ????܂??B
?@???Ђ??}??`???Ƃ??ɂ́A?Ђ??̏???Ɖ????????̂܂܍ő?l?ƍŏ??l?ɂ͂????ɁA?u??3?l???ʐ??|??1?l???ʐ??v?̒l??1.5?{?????l???3?l???ʐ??ɉ??Z?A???邢?͑?1?l???ʐ????猸?Z???????̂?????Ɖ????ɂ??邱?Ƃ?????܂??i??3?l???ʐ??|??1?l???ʐ??̒l???u?l???ʔ͈́v?ƌĂԂ??Ƃ?????܂??j?BMedInc??͂??̂悤?Ȍ`?ŕ`???ꂽ???Ђ??}?ł??BMedInc??ł́u??3?l???ʐ??|??1?l???ʐ??v?͂????悻?u2.18?v?ł??B?????1.5?{?????l?i3.27?j???A??1?l???ʐ????猸?Z?A??3?l???ʐ??ɉ??Z???????̂??ȉ??Ɏ????܂??B
- ?????F2.5634?|3.27???|0.706?i?ŏ??l??0.4999?j
- ??1?l???ʐ??F2.5634
- ?????l?F3.5348
- ??3?l???ʐ??F4.74325
- ????F4.74325?{3.27??8.01?i?ő?l??15.0001?j
?@?????̒l??MedInc??̔??Ђ??}??????ׂ?Ɖ????͎??ۂ̍ŏ??l???????????Ȃ??Ă???̂ł??̂܂܍ŏ??l?̕????ɉ??????????A????͍ő?l?܂œ͂??Ȃ??̂ŏ???̕ӂ?ʼn??????????????̂ƂȂ??Ă??邱?Ƃ???????܂??B
?????A??????l?????????Ђ??}?@????≺??????O?ꂽ?l?́u?O??l?v?i???炩?̌????ő??̃f?[?^?Ƃ͑傫???قȂ?l?ƂȂ????f?[?^?j?Ƃ??Ĉ????邱?Ƃ??悭????܂????A???̗?ł?????????????????̂??K???ǂ????͔????ȂƂ???ł??B?????????f?[?^?????߂??邩??ł??B????͍??????ґw?̃f?[?^???A?v?Z???ꂽ??????Ċɂ₩?ɕ??z???Ă???ƍl?????????悢????????܂???B
?@???ɂ??̍l?????????Ă??邩?ǂ??????A?q?X?g?O???????g???Č??Ă݂܂??傤?B?Ȃ??A?ȉ??̃R?[?h?????s????ƁA?ǂݍ??f?[?^?̑S?Ă̗?ɂ??āA???Ђ??}???쐬???A1?̃O???t?̒??ɕ\???ł??܂??B?????̂?????͎????Ă݂Ă????????B
fig = plt.figure()
fig, axes = plt.subplots(nrows=3, ncols=3, sharex=False)
for k, v in col.items():
r = v // 3
c = v % 3
axes[r, c].boxplot(data[:, v])
axes[r, c].set_title(k, fontsize=8)
plt.show()
Matplotlib?ɂ??q?X?g?O?????̕`??
?@?q?X?g?O???????܂??f?[?^?̕??z?̎d???ׂ邽?߂̃c?[???ł??B??????̓f?[?^?Z?b?g??????͈̔͂ɕ??????āA???͈̔͂Ɋ܂܂??f?[?^???????邩?ׂ܂??i?ł??????̂?x?????z?\?Ƃ????܂??j?B?????ł?MedInc????Ɏ??܂????A?C???[?W?Ƃ??Ă͏????????͈̔͂̐l?????l?A???͈̔͂̐l?????l?Ƃ????????Ƃׂ?Ƃ??????Ƃł??B
?@?x???̕??z?????߂?ɂ?NumPy??numpy.histogram?????g???A????ꂽ???ʂ?matplotlib.pyplot.bar???ŃO???t?ɂ???Ƃ????̂??傴???ςȎ菇?ɂȂ?܂??B
?@?܂?numpy.histogram?????猩?Ă????܂??傤?B
numpy.histogram(a, bins=10)
?@???ɂ??p?????[?^?[?͂???܂????A?????ł?2?????p?????[?^?[???w?肵?܂??傤?B
- a?F???͂???f?[?^
- bins?F?f?[?^?͈͂????̗̈?ɕ??????邩?̎w??B?f?t?H???g?l??10
?@???͂???f?[?^?ƃr???i?f?[?^?Z?b?g???????X?̋?ԁj?̐????w?肵?āA???̊????Ăяo???ƁA?q?X?g?O?????i?e?r???Ɋ܂܂??f?[?^?̐??j?A???ꂩ??r???ƃr???̋??E?ƂȂ?l???܂z?Ԃ???܂??B
hist, bins = np.histogram(data[:, col['MedInc']], bins=20)
?@???̗?ł̓f?[?^?Z?b?g?Ɋ܂܂??l?͈̔́i?ŏ??l?ł???0.4999????ő?l?ł???15.0001?j??20?̃r???ɕ??????āA???ꂼ??̃r???Ɋ܂܂??f?[?^?̐?????уr???ƃr???̋??E?ƂȂ?l????̂ł??B
?@?ȉ??́A???̃R?[?h?????s???āAhist??bins??2?̕ϐ??ɕԂ??ꂽ?l??\???????Ƃ???ł??B
?@?ϐ?hist?̒l?????????A???????Ⴂ?????璆???炢?ɂ????Ă͂??Ȃ?̃{?????[????????܂????A???????ґw?ɂ??Ă?????Ȃ?̐??????邱?Ƃ??m???ɕ?????܂??B??̔??Ђ??}?ɂ????錟???ł??????O??l?Ƃ??ׂ??ł͂Ȃ?????????Ȃ??Ƃ????l???͂?????x???????Ƃ?????ł??傤?B
?@???̉??̕ϐ?bins?̒l?͊e?r???̍ŏ??l?ƍő?l??\?????E???ƍl?????܂??B?܂?A?ŏ??̃r????0.4999????1.22491?܂ł͈̔͂Ɋ܂܂??f?[?^?̐???\???Ă???Ƃ??????Ƃł??i???̂??߁Abins?̗v?f????hist?̗v?f??????1?????Ȃ??Ă??܂??j?B
?@??͂?????O???t?Ƃ??ĕ`?悷?邾???ł??B??قǂ????????悤?ɁA????ɂ?matplotlib.pyplot.bar?????g???܂??B
plt.bar(bins[0:-1], hist, align='edge')
plt.xticks(bins, fontsize=6)
plt.show()
?@bar???̑?1??????X???̍??W???w?肵?܂??B?????ł͒P??histogram?????Ԃ??Ă??????E?l???܂ޔz??Ō??1????菜???????̂ɂ??܂????B??2?????ɂ?Y???̒l?A?܂?A?e?r???̃f?[?^?????w?肵?܂??B??̃R?[?h?ł?align?p?????[?^?[??'edge'??n???Ă??܂????A????ɂ??O???t??X???̖ڐ???ɑ??č?????܂??B
?@????X???̖ڐ????ݒ肵?Ă???̂????̍s?ł??B?????ł?X???ɍŏ??l????ő?l?܂ł??\???????悤?ɁAhistogram?????Ԃ??Ă??????E?l??n???Ă??܂??B?܂??A?t?H???g?T?C?Y???????߂ɂ??āA?S?Ă̒l???Ȃ?ׂ??\???ł???悤?ɂ??܂????B
?@?ȉ??????s???ʂł??B
???s?????@?ǂ??ł??傤?B?????̗???????Ă??邾???????A?????ґw?̍\???????n?b?L???Ɣc???ł???̂ł͂Ȃ??ł??傤???B???????҂??ɂ₩?ɕ??z???Ă??邾?낤?Ƃ????????????Ȃ????ԈႢ?ł͂Ȃ??????ł??B
?@?????????X???́A??{???v?ʂ??????????x?͐????ł??邩??????܂???B?Ⴆ?AMedInc??Ɋւ??ẮA?O??̍Ō?Ɏ??̂悤?Ɍ??????Ă??܂??B
- MedInc?F???ς?3.87???x?A?W??????1.9???x?ł??邱?Ƃ??珊????2?`6???h???̊ԂɍL?????Ă???Ǝv????B???̈???ŁA?ő?l??15.0?Ƃ??Ȃ?傫?Ȓl?ɂȂ??Ă???B???Ȃ?̍??????Ă???l?????i?u???b?N?O???[?v?j?????݂??邩??????Ȃ?
?@???̌????ƃO???t?Ƃ?????ׂĂ݂?ƁA?܂??܂?????Ȃ?̃f?L???????ƕM?ғI?ɂ͊????Ă??܂??܂??i?Â??ł????ˁj?B
?@??{???v?ʂ????ł?????Ȃ?̌??????ł??????ŁA????Љ?????Ђ??}??q?X?g?O???????g???Č??????????邱?ƂŁA?茳?ɂ???f?[?^?Z?b?g?ɑ??闝??????i?Ɛ[?܂?͂??ł??i?????ł̈?ۂ??????ɂ???ĕ⋭????邩??????܂??A?t?Ɉ?ۂ??܂???ƂЂ?????Ԃ邱?ƂɂȂ邩??????܂???j?B
?@???̂悤?ɂ??܂??܂ȃc?[????p???ăf?[?^?Z?b?g????舵?????ƂŁA???܂?ɖc??ŌX?̃f?[?^???????ł??Ȃ??i???????Ă????܂?Ӗ????Ȃ??j?f?[?^?Z?b?g????????̓??????????Ă??܂??B
?@?Ƃ????킯?ŁA????͔??Ђ??}?ƃq?X?g?O?????ɂ??Č??Ă݂܂????B????͂??̂悤?ȉ????ɂ??Ă??????????Ă݂邩?A?܂??ʂ̂??b?ɂ??邩?A?Y??ł???Ƃ???ł??B
Copyright© Digital Advantage Corp. All Rights Reserved.
?A?C?e?B???f?B?A????̂??m?点
??IT eBook
???ڂ̃e?[?}
?ҏW??????̂??m?点
RSS?ɂ???
?A?C?e?B???f?B?AID?ɂ???
???[???}?K?W???o?^
??IT?̃??[???}?K?W???́A ???????A???ׂĖ????ł??B???Ѓ??[???}?K?W???????w?ǂ????????B
ITmedia?̓A?C?e?B???f?B?A??????Ђ̓o?^???W?ł??B
???f?B?A?ꗗ | ????SNS | ?L???ē? | ???₢???킹 | ?v???C?o?V?[?|???V?[ | RSS | ?^?c??? | ?̗p??? | ??????