データ分析コンペにおいて 特徴量管理に疲弊している全人類に伝えたい想い

Transcript

1. 1.

   ϚϚͷҰาΛࢧ͑Δ σʔλ෼ੳίϯϖʹ͓͍ͯ
   ಛ௃ྔ؅ཧʹർฐ͍ͯ͠Δશਓྨʹ఻͍͑ͨ૝͍
   ʙֶशɾਪ࿦ύΠϓϥΠϯΛఴ͑ͯʙ $POOFIJUP
   *OD
   ໺ᖒ఩র
   
   $POOFIJUP
   .BSDIÉ
   WPM
   ʙػցֶशɾσʔλ෼ੳࢢʙ

2. 2.

   ͜Μʹͪ͸ʂ ϚϚͷҰาΛࢧ͑Δ

3. 3.

   ͍͖ͳΓͰ͕͢ ϚϚͷҰาΛࢧ͑Δ

4. 4.

   σʔλ෼ੳίϯϖʢ,BHHMF
   
   4*(/"5&ͳͲʣ
   ฉ͍ͨ͜ͱ͋Δਓʙ! ϚϚͷҰาΛࢧ͑Δ

5. 5.

   σʔλ෼ੳίϯϖʹࢀՃͨ͜͠ͱ͋Δਓʙ! ϚϚͷҰาΛࢧ͑Δ

6. 6.

   ಛ௃ྔͷ؅ཧͬͯͲ͏ͯ͠·͔͢ʁ
   ʢςʔϒϧσʔλʹ͓͍ͯʣ ϚϚͷҰาΛࢧ͑Δ

7. 7.

   Α͋͘Δύλʔϯʢ࣮ମݧʣ ϚϚͷҰาΛࢧ͑Δ

8. 8.

   Α͋͘ΔύλʔϯʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ w ಛ௃ྔ࡞Δ
   DPM
   
   <
   Z
   "
   #
   $
   %
   >
   
   ˠ
   
   <
   Z
   "

   
   #
   $
   %
   &
   '
   
   > # e.g) train['A'] = train['A'].fillna(0) train['B'] = np.log1p(train['B']) train['E'] = train['A'] + train['B'] df_group = train.groupby('D')['E'].mean() train['F'] = train['D'].map(df_group) <> ɾ
   ɾ
   ɾ
9. 9.

   Α͋͘ΔύλʔϯʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ w ࢖͏ಛ௃ྔͷΧϥϜ͚ͩࢦఆ͢Δ # e.g) feat_col = ['A', 'C',

   'D', 'E', 'F', 'J'] x_train = train[feat_col] y_train = train['y'] # e.g) clf.fit(x_train, y_train) w ֶशͤ͞Δ <> <> ɾ
   ɾ
   ɾ
10. 10.

    Α͋͘ΔύλʔϯʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ w ࢖͏ಛ௃ྔͷΧϥϜ͚ͩࢦఆ͢Δ # e.g) feat_col = ['A', 'C',

    'D', 'E', 'F', 'J'] x_train = train[feat_col] y_train = train['y'] # e.g) clf.fit(x_train, y_train) w ֶशͤ͞Δ <> <> ɾ
    ɾ
    ɾ ͋Ε
    b'`ͬͯͲΜͳಛ௃ྔ͚ͩͬʁ
11. 11.

    Α͋͘ΔύλʔϯʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ # e.g) train['A'] = train['A'].fillna(0) train['B'] = np.log1p(train['B'])

    train['E'] = train['A'] + train['B'] df_group = train.groupby('D')['E'].mean() train['F'] = train['D'].map(df_group) <> # e.g) feat_col = ['A', 'C', 'D', 'E', 'F', 'J'] x_train = train[feat_col] y_train = train['y'] # e.g) clf.fit(x_train, y_train) <> <>
12. 12.

    Α͋͘ΔύλʔϯʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ # e.g) train['A'] = train['A'].fillna(0) train['B'] = np.log1p(train['B'])

    train['E'] = train['A'] + train['B'] df_group = train.groupby('D')['E'].mean() train['F'] = train['D'].map(df_group) <> # e.g) feat_col = ['A', 'C', 'D', 'E', 'F', 'J'] x_train = train[feat_col] y_train = train['y'] # e.g) clf.fit(x_train, y_train) <> <> ݟ͚ͭͨʂ
    ʢOPUFCPPLͷ্ͷํʣ
13. 13.

    Α͋͘ΔύλʔϯʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ # e.g) train['A'] = train['A'].fillna(0) train['B'] = np.log1p(train['B'])

    train['E'] = train['A'] + train['B'] df_group = train.groupby('D')['E'].mean() train['F'] = train['D'].map(df_group) <> # e.g) feat_col = ['A', 'C', 'D', 'E', 'F', 'J'] x_train = train[feat_col] y_train = train['y'] # e.g) clf.fit(x_train, y_train) <> <> ݟ͚ͭͨʂ
    ʢOPUFCPPLͷ্ͷํʣ ಛ௃ྔ͕গͳ͍৔߹͸·ͩϚγ͕ͩɺ
    ଟ͘ͳͬͯ͘ΔͱͲΜͳܭࢉͰٻΊͨ
    ಛ௃ྔ͔ͩͬͨΛ͍͍ͪͪߟ͑Δʢ୳͢ʣ
    ͷ͸݁ߏେมͩ͠ɺ͕͔͔࣌ؒΔ
14. 14.

    Α͋͘Δύλʔϯ
    ͦͷʢ࣮ମݧʣ ϚϚͷҰาΛࢧ͑Δ

15. 15.

    Α͋͘Δύλʔϯ
    ͦͷʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ Αͬ͠Ό͊ʂΊͬͪΌྑ͍είΞͰͨͥʙʙʙ
    ͜ͷOPUFCPPLΛ%VQMJDBUFͯ͠ɺ΋ͬͱྑ͍Ϟσϧ࡞ͬͪΌ͏ͧʂ

16. 16.

    ҰํɺOPUFCPPLͷத਎͸ʜ ϚϚͷҰาΛࢧ͑Δ

17. 17.

    Α͋͘Δύλʔϯ
    ͦͷʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ <> import numpy as np import pandas as

    pd OPUFCPPLͷத਎ ɾ
    ɾ
    ɾ <> submission.to_csv('submission.csv', index=False)
18. 18.

    Α͋͘Δύλʔϯ
    ͦͷʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ <> import numpy as np import pandas as

    pd OPUFCPPLͷத਎ ɾ
    ɾ
    ɾ ɾ
    ɾ
    ɾ <> submission.to_csv('submission.csv', index=False)
19. 19.

    Α͋͘Δύλʔϯ
    ͦͷʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ <> import numpy as np import pandas as

    pd OPUFCPPLͷத਎ ɾ
    ɾ
    ɾ ɾ
    ɾ
    ɾ <> submission.to_csv('submission.csv', index=False)
20. 20.

    Α͋͘Δύλʔϯ
    ͦͷʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ <> import numpy as np import pandas as

    pd OPUFCPPLͷத਎ ɾ
    ɾ
    ɾ ɾ
    ɾ
    ɾ <> submission.to_csv('submission.csv', index=False) ಉ͡ܭࢉΛԿ౓΋΍Βͳ͍ͱ͍͚ͳ͍ ʴ ºʢ ʣˠ
    ແବ
21. 21.

    Α͋͘Δύλʔϯ
    ͦͷʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ ౓ॏͳΔ%VQMJDBUFʹΑΓɺOPUFCPPL஍ࠈʹؕΔՄೳੑ΋ʜ dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    

    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    ʜʜʜʜʜ
22. 22.

    Α͋͘Δύλʔϯ
    ͦͷʢJQZOCʣ ϚϚͷҰาΛࢧ͑Δ ౓ॏͳΔ%VQMJDBUFʹΑΓɺOPUFCPPL஍ࠈʹؕΔՄೳੑ΋ʜ dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    

    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    dOPUFCPPL-JHIU#(.@TDPSF@JQZOC
    ʜʜʜʜʜ
23. 23.

    ϚϚͷҰาΛࢧ͑Δ ࠓ೔࿩͢͜ͱ

24. 24.

    ϚϚͷҰาΛࢧ͑Δ σʔλ෼ੳίϯϖʹ͓͍ͯ
    ಛ௃ྔ؅ཧʹർฐ͍ͯ͠Δશਓྨʹ఻͍͑ͨ૝͍
    ʙֶशɾਪ࿦ύΠϓϥΠϯΛఴ͑ͯʙ

25. 25.

    ΞδΣϯμ  ࣗݾ঺հ
     ಛ௃ྔ؅ཧʹ͍ͭͯ
    ‣ ྻ͝ͱʹQJDLMFϑΝΠϧͰಛ௃ྔΛ؅ཧ
    ‣ ಛ௃ྔੜ੒࣌ɺಉ࣌ʹϝϞϑΝΠϧ΋ੜ੒
    

    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ
    ‣ ίϚϯυҰൃͰֶशˠ4VCNJUϑΝΠϧ࡞੒·ͰΛ࣮ߦ
    ‣ ֶशʹ࢖༻ͨ͠ಛ௃ྔ΍Ϟσϧύϥϝʔλ͸MPHͱҰॹʹอଘ
    ‣ TIBQΛ༻͍ͯಛ௃ྔͷߩݙ౓ΛՄࢹԽ͠ɺ࣍ճֶश࣌ͷצॴΛݟ͚ͭΔ ϚϚͷҰาΛࢧ͑Δ
26. 26.

    ΞδΣϯμ  ࣗݾ঺հ
     ಛ௃ྔ؅ཧʹ͍ͭͯ
    ‣ ྻ͝ͱʹQJDLMFϑΝΠϧͰಛ௃ྔΛ؅ཧ
    ‣ ಛ௃ྔੜ੒࣌ɺಉ࣌ʹϝϞϑΝΠϧ΋ੜ੒
    

    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ
    ‣ ίϚϯυҰൃͰֶशˠ4VCNJUϑΝΠϧ࡞੒·ͰΛ࣮ߦ
    ‣ ֶशʹ࢖༻ͨ͠ಛ௃ྔ΍Ϟσϧύϥϝʔλ͸MPHͱҰॹʹอଘ
    ‣ TIBQΛ༻͍ͯಛ௃ྔͷߩݙ౓ΛՄࢹԽ͠ɺ࣍ճֶश࣌ͷצॴΛݟ͚ͭΔ ϚϚͷҰาΛࢧ͑Δ ݰਓͷ஌ܙΛ͓आΓͨ͠Β
    ΊͬͪΌΑ͔ͬͨ ʢ˞ʣ ͍ͬͯ͏࿩Λ͠·͢
    ʢ˞ʣ͋͘·Ͱओ؍Ͱ͢
27. 27.

    
    ࣗݾ঺հ ϚϚͷҰาΛࢧ͑Δ

28. 28.

    
    ࣗݾ঺հ ϚϚͷҰาΛࢧ͑Δ ໊લɿ໺ᖒ
    ఩রʢ/P[BXB
    5BLBOPCVʣ
    ॴଐɿίωώτגࣜձࣾ
    ɹɹɿ͔ͨͺ͍!UBLBQZ
    w ʙίωώτʹ.-ΤϯδχΞͱͯ͠+0*/
    w ػցֶशʢ/-1ɺਪનγεςϜʣΛϝΠϯʹ΍ΓͭͭΠϯϑϥʢ"84ʣ΋ษڧத
    w

    ,BHHMFͨ͠ΓɺϒϩάʢIUUQTXXXUBLBQZXPSLʣॻ͍ͨΓɺ໺ٿͨ͠Γɺ
    
    
    ϥʔϝϯ৯΂ͨΓ͍ͯ͠·͢
29. 29.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ˞ԼههࣄΛࢀߟʹ͍͖ͤͯͨͩ͞·ͨ͠ɻ
    ɾ,BHHMFͰ࢖͑Δ'FBUIFSܗࣜΛར༻ͨ͠ಛ௃ྔ؅ཧ๏
    
    
    
    IUUQTBNBMPHIBUFCMPKQFOUSZLBHHMFGFBUVSFNBOBHFNFOU ‣ ྻ͝ͱʹQJDLMFϑΝΠϧͰಛ௃ྔΛ؅ཧ
    ‣ ಛ௃ྔੜ੒࣌ɺಉ࣌ʹϝϞϑΝΠϧ΋ੜ੒

30. 30.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ˞ԼههࣄΛࢀߟʹ͍͖ͤͯͨͩ͞·ͨ͠ɻ
    ɾ,BHHMFͰ࢖͑Δ'FBUIFSܗࣜΛར༻ͨ͠ಛ௃ྔ؅ཧ๏
    
    
    
    IUUQTBNBMPHIBUFCMPKQFOUSZLBHHMFGFBUVSFNBOBHFNFOU ‣ ྻ͝ͱʹQJDLMFϑΝΠϧͰಛ௃ྔΛ؅ཧ
    ‣ ಛ௃ྔੜ੒࣌ɺಉ࣌ʹϝϞϑΝΠϧ΋ੜ੒ ࠷ॳʹΠϝʔδΛڞ༗͠·͢

31. 31.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ lྻ͝ͱzʹಛ௃ྔΛQJDLMFϑΝΠϧͰ؅ཧ͢Δ

32. 32.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ 4VSWJWFE 1DMBTT 4FY "HF &NCBSLFE   NBMF
    

     4   NBMF  $   GFNBMF  $   NBMF  $   GFNBMF  $   GFNBMF  4   NBMF  4 lྻ͝ͱzʹಛ௃ྔΛQJDLMFϑΝΠϧͰ؅ཧ͢Δ
33. 33.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ 4VSWJWFE 1DMBTT 4FY "HF &NCBSLFE   NBMF
    

     4   NBMF  $   GFNBMF  $   NBMF  $   GFNBMF  $   GFNBMF  4   NBMF  4 TVSWJWFE@USBJOQLM QDMBTT@USBJOQLM
    QDMBTT@UFTUQLM TFY@USBJOQLM
    TFY@UFTUQLM BHF@USBJOQLM
    BHF@UFTUQLM FNCBSLFE@USBJOQLM
    FNCBSLFE@UFTUQLM lྻ͝ͱzʹಛ௃ྔΛQJDLMFϑΝΠϧͰ؅ཧ͢Δ
34. 34.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ಛ௃ྔੜ੒࣌ɺಉ࣌ʹಛ௃ྔϝϞΛ࡞੒͢Δ

35. 35.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ಛ௃ྔੜ੒࣌ɺಉ࣌ʹಛ௃ྔϝϞΛ࡞੒͢Δ

36. 36.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ಛ௃ྔੜ੒࣌ɺಉ࣌ʹಛ௃ྔϝϞΛ࡞੒͢Δ ݁ߏେมͦ͏ɾɾɾ

37. 37.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ಛ௃ྔੜ੒࣌ɺಉ࣌ʹಛ௃ྔϝϞΛ࡞੒͢Δ Ͱ΋ɺQZUIPOεΫϦϓτΛ࣮ͭߦ͢Δ͚ͩɻ

38. 38.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ IPHFQZ
    ΛίϚϯυϥΠϯ͔Β࣮ߦ͢Δ͚ͩ class Pclass(Feature): def create_features(self): self.train['Pclass'] = train['Pclass']

    self.test['Pclass'] = test['Pclass'] create_memo('Pclass','νέοτͷΫϥεɻ1st, 2nd, 3rdͷ3छྨ') class Sex(Feature): def create_features(self): self.train['Sex'] = train['Sex'] self.test['Sex'] = test['Sex'] create_memo('Sex','ੑผ') class Age(Feature): def create_features(self): self.train['Age'] = train['Age'] self.test['Age'] = test['Age'] create_memo('Age','೥ྸ') class Age_mis_val_median(Feature): def create_features(self): self.train['Age_mis_val_median'] = train['Age'].fillna(train['Age'].median()) self.test['Age_mis_val_median'] = test['Age'].fillna(test['Age'].median()) create_memo('Age_mis_val_median','೥ྸͷܽଛ஋Λதԝ஋Ͱิ׬ͨ͠΋ͷ')
39. 39.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Pclass(Feature): def create_features(self): self.train['Pclass'] = train['Pclass'] self.test['Pclass']

    = test['Pclass'] create_memo('Pclass','νέοτͷΫϥεɻ1st, 2nd, 3rdͷ3छྨ') class Sex(Feature): def create_features(self): self.train['Sex'] = train['Sex'] self.test['Sex'] = test['Sex'] create_memo('Sex','ੑผ') class Age(Feature): def create_features(self): self.train['Age'] = train['Age'] self.test['Age'] = test['Age'] create_memo('Age','೥ྸ') class Age_mis_val_median(Feature): def create_features(self): self.train['Age_mis_val_median'] = train['Age'].fillna(train['Age'].median()) self.test['Age_mis_val_median'] = test['Age'].fillna(test['Age'].median()) create_memo('Age_mis_val_median','೥ྸͷܽଛ஋Λதԝ஋Ͱิ׬ͨ͠΋ͷ') IPHFQZ
    ΛίϚϯυϥΠϯ͔Β࣮ߦ͢Δ͚ͩ
40. 40.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Pclass(Feature): def create_features(self): self.train['Pclass'] = train['Pclass'] self.test['Pclass']

    = test['Pclass'] create_memo('Pclass','νέοτͷΫϥεɻ1st, 2nd, 3rdͷ3छྨ') class Sex(Feature): def create_features(self): self.train['Sex'] = train['Sex'] self.test['Sex'] = test['Sex'] create_memo('Sex','ੑผ') class Age(Feature): def create_features(self): self.train['Age'] = train['Age'] self.test['Age'] = test['Age'] create_memo('Age','೥ྸ') class Age_mis_val_median(Feature): def create_features(self): self.train['Age_mis_val_median'] = train['Age'].fillna(train['Age'].median()) self.test['Age_mis_val_median'] = test['Age'].fillna(test['Age'].median()) create_memo('Age_mis_val_median','೥ྸͷܽଛ஋Λதԝ஋Ͱิ׬ͨ͠΋ͷ') ֤ಛ௃ྔ IPHFQZ
    ΛίϚϯυϥΠϯ͔Β࣮ߦ͢Δ͚ͩ
41. 41.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Pclass(Feature): def create_features(self): self.train['Pclass'] = train['Pclass'] self.test['Pclass']

    = test['Pclass'] create_memo('Pclass','νέοτͷΫϥεɻ1st, 2nd, 3rdͷ3छྨ') class Sex(Feature): def create_features(self): self.train['Sex'] = train['Sex'] self.test['Sex'] = test['Sex'] create_memo('Sex','ੑผ') class Age(Feature): def create_features(self): self.train['Age'] = train['Age'] self.test['Age'] = test['Age'] create_memo('Age','೥ྸ') class Age_mis_val_median(Feature): def create_features(self): self.train['Age_mis_val_median'] = train['Age'].fillna(train['Age'].median()) self.test['Age_mis_val_median'] = test['Age'].fillna(test['Age'].median()) create_memo('Age_mis_val_median','೥ྸͷܽଛ஋Λதԝ஋Ͱิ׬ͨ͠΋ͷ') ಛ௃ྔϝϞϑΝΠϧ IPHFQZ
    ΛίϚϯυϥΠϯ͔Β࣮ߦ͢Δ͚ͩ
42. 42.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Pclass(Feature): def create_features(self): self.train['Pclass'] = train['Pclass'] self.test['Pclass']

    = test['Pclass'] create_memo('Pclass','νέοτͷΫϥεɻ1st, 2nd, 3rdͷ3छྨ') DSFBUF@NFNPͷॲཧ֓ཁ
43. 43.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Pclass(Feature): def create_features(self): self.train['Pclass'] = train['Pclass'] self.test['Pclass']

    = test['Pclass'] create_memo('Pclass','νέοτͷΫϥεɻ1st, 2nd, 3rdͷ3छྨ') DSFBUF@NFNPͷॲཧ֓ཁ # ಛ௃ྔϝϞcsvϑΝΠϧ࡞੒ def create_memo(col_name, desc): file_path = Feature.dir + '/_features_memo.csv' if not os.path.isfile(file_path): with open(file_path,"w"):pass with open(file_path, 'r+') as f: lines = f.readlines() lines = [line.strip() for line in lines] # ॻ͖ࠐ΋͏ͱ͍ͯ͠Δಛ௃ྔ͕͢Ͱʹॻ͖ࠐ·Ε͍ͯͳ͍͔νΣοΫ col = [line for line in lines if line.split(',')[0] == col_name] if len(col) != 0:return writer = csv.writer(f) writer.writerow([col_name, desc])
44. 44.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Pclass(Feature): def create_features(self): self.train['Pclass'] = train['Pclass'] self.test['Pclass']

    = test['Pclass'] create_memo('Pclass','νέοτͷΫϥεɻ1st, 2nd, 3rdͷ3छྨ') DSFBUF@NFNPͷॲཧ֓ཁ # ಛ௃ྔϝϞcsvϑΝΠϧ࡞੒ def create_memo(col_name, desc): file_path = Feature.dir + '/_features_memo.csv' if not os.path.isfile(file_path): with open(file_path,"w"):pass with open(file_path, 'r+') as f: lines = f.readlines() lines = [line.strip() for line in lines] # ॻ͖ࠐ΋͏ͱ͍ͯ͠Δಛ௃ྔ͕͢Ͱʹॻ͖ࠐ·Ε͍ͯͳ͍͔νΣοΫ col = [line for line in lines if line.split(',')[0] == col_name] if len(col) != 0:return writer = csv.writer(f) writer.writerow([col_name, desc])
45. 45.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Pclass(Feature): def create_features(self): self.train['Pclass'] = train['Pclass'] self.test['Pclass']

    = test['Pclass'] create_memo('Pclass','νέοτͷΫϥεɻ1st, 2nd, 3rdͷ3छྨ') DSFBUF@NFNPͷॲཧ֓ཁ # ಛ௃ྔϝϞcsvϑΝΠϧ࡞੒ def create_memo(col_name, desc): file_path = Feature.dir + '/_features_memo.csv' if not os.path.isfile(file_path): with open(file_path,"w"):pass with open(file_path, 'r+') as f: lines = f.readlines() lines = [line.strip() for line in lines] # ॻ͖ࠐ΋͏ͱ͍ͯ͠Δಛ௃ྔ͕͢Ͱʹॻ͖ࠐ·Ε͍ͯͳ͍͔νΣοΫ col = [line for line in lines if line.split(',')[0] == col_name] if len(col) != 0:return writer = csv.writer(f) writer.writerow([col_name, desc]) $47ܗࣜͰอଘ͓ͯ͘͠ͱ(JUIVC͔Βࢀর͠΍͍͢
    ʢ΋ͪΖΜɺ&YDFM΍/VNCFSTͱ͍ͬͨΞϓϦέʔγϣϯ͔ΒͰ΋៉ྷʹݟ͑Δʣ
46. 46.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ৽͍͠ಛ௃ྔΛ࡞੒͢Δ৔߹

47. 47.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Family_Size(Feature): def create_features(self): self.train['Family_Size'] = train['Parch'] +

    train['SibSp'] self.test['Family_Size'] = test['Parch'] + test['SibSp'] create_memo('Family_Size','Ո଒ͷ૯਺') IPHFQZ
    ʹ৽͍͠ಛ௃ྔੜ੒ॲཧΛهड़ ৽͍͠ಛ௃ྔΛ࡞੒͢Δ৔߹
48. 48.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Family_Size(Feature): def create_features(self): self.train['Family_Size'] = train['Parch'] +

    train['SibSp'] self.test['Family_Size'] = test['Parch'] + test['SibSp'] create_memo('Family_Size','Ո଒ͷ૯਺') 
    QZUIPO
    IPHFQZ ৽͍͠ಛ௃ྔΛ࡞੒͢Δ৔߹ IPHFQZ
    ʹ৽͍͠ಛ௃ྔੜ੒ॲཧΛهड़
49. 49.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ class Family_Size(Feature): def create_features(self): self.train['Family_Size'] = train['Parch'] +

    train['SibSp'] self.test['Family_Size'] = test['Parch'] + test['SibSp'] create_memo('Family_Size','Ո଒ͷ૯਺') 
    QZUIPO
    IPHFQZ ৽͍͠ಛ௃ྔͷΈੜ੒ ৽͍͠ಛ௃ྔΛ࡞੒͢Δ৔߹ IPHFQZ
    ʹ৽͍͠ಛ௃ྔੜ੒ॲཧΛهड़
50. 50.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ σʔλΛಡΈࠐΉࡍ͸ɺಛ௃ྔΛࢦఆͯ͠ϩʔυ͢Δ͚ͩ # ಛ௃ྔͷࢦఆ features = [ "age_mis_val_median", "family__size",

    "cabin", "fare_mis_val_median" ] df = [pd.read_pickle(FEATURE_DIR_NAME + f’{f}_train.pkl') for f in features] df = pd.concat(df, axis=1)
51. 51.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ σʔλΛಡΈࠐΉࡍ͸ɺಛ௃ྔΛࢦఆͯ͠ϩʔυ͢Δ͚ͩ # ಛ௃ྔͷࢦఆ features = [ "age_mis_val_median", "family__size",

    "cabin", "fare_mis_val_median" ] df = [pd.read_pickle(FEATURE_DIR_NAME + f’{f}_train.pkl') for f in features] df = pd.concat(df, axis=1) Կ͕خ͔͔ͬͨ͠
52. 52.

    
    ಛ௃ྔ؅ཧʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ w ͭͷεΫϦϓτϑΝΠϧʹಛ௃ྔੜ੒Λ·ͱΊΔ͜ͱͰɺಉ͡ܭࢉΛෳ਺ ճ࣮ߦ͢Δ͜ͱΛආ͚ɺ࣌ؒΛ༗ޮ׆༻Ͱ͖Δɻ
    ɹˠಛ௃ྔͷ࠶ݱੑ΋୲อɻ
    w ಛ௃ྔͷϝϞΛಉ࣌ʹੜ੒͢Δ͜ͱͰʮ͜ͷಛ௃ྔͳΜ͚ͩͬʁʯͱ಄Λ࢖ ΘͣʹࡁΜͩɻ
    w

    ಛ௃ྔΛྻ͝ͱʹ؅ཧ͢Δ͜ͱͰऔΓճָ͕͠ʹͳΔɻ
    ɹˠQJDLMFϑΝΠϧͩͱอଘ΋ಡΈࠐΈ΋଎͍ʂ
    ɹˠಛ௃ྔ͕๲େʹͳΔ৔߹͸ɺ͋Δఔ౓ͷ୯ҐͰ·ͱΊͯ؅ཧ͢Δํ͕ྑ͍͔΋ɻ
53. 53.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ˞Լهॻ੶ʹܝࡌ͞Ε͍ͯΔύΠϓϥΠϯΛࢀߟʹ͠·ͨ͠ɻ
    ɾ,BHHMFͰউͭσʔλ෼ੳͷٕज़
    
    
    
    IUUQTHJIZPKQCPPL ‣ ίϚϯυҰൃͰֶशˠ4VCNJUϑΝΠϧ࡞੒·ͰΛ࣮ߦ
    ‣ ֶशʹ࢖༻ͨ͠ಛ௃ྔ΍Ϟσϧύϥϝʔλ͸MPHͱҰॹʹอଘ
    ‣

    TIBQΛ༻͍ͯಛ௃ྔͷߩݙ౓ΛՄࢹԽ͠ɺ࣍ճֶश࣌ͷצॴΛݟ͚ͭΔ
54. 54.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ SVOQZ
    Λ࣮ߦ͢Δ͜ͱͰɺֶशɾਪ࿦ɾ4VCNJUϑΝΠϧΛ࡞੒ # ಛ௃ྔͷࢦఆ features = [ "age_mis_val_median", "family__size",

    "cabin", "fare_mis_val_median" ] run_name = 'lgb_1102' # ࢖༻͢Δಛ௃ྔϦετͷอଘ with open(LOG_DIR_NAME + run_name + "_features.txt", 'wt') as f: for ele in features: f.write(ele+'\n') params_lgb = { 'boosting_type': 'gbdt', 'objective': 'binary', 'early_stopping_rounds': 20, 'verbose': 10, 'random_state': 99, 'num_round': 100 } # ࢖༻͢Δύϥϝʔλͷอଘ with open(LOG_DIR_NAME + run_name + "_param.txt", 'wt') as f: for key,value in sorted(params_lgb.items()): f.write(f'{key}:{value}\n') runner = Runner(run_name, ModelLGB, features, params_lgb, n_fold, name_prefix) runner.run_train_cv() # ֶश runner.run_predict_cv() # ਪ࿦ Submission.create_submission(run_name) # submit࡞੒
55. 55.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ SVOQZ
    Λ࣮ߦ͢Δ͜ͱͰɺֶशɾਪ࿦ɾ4VCNJUϑΝΠϧΛ࡞੒ # ಛ௃ྔͷࢦఆ features = [ "age_mis_val_median", "family__size",

    "cabin", "fare_mis_val_median" ] run_name = 'lgb_1102' # ࢖༻͢Δಛ௃ྔϦετͷอଘ with open(LOG_DIR_NAME + run_name + "_features.txt", 'wt') as f: for ele in features: f.write(ele+'\n') params_lgb = { 'boosting_type': 'gbdt', 'objective': 'binary', 'early_stopping_rounds': 20, 'verbose': 10, 'random_state': 99, 'num_round': 100 } # ࢖༻͢Δύϥϝʔλͷอଘ with open(LOG_DIR_NAME + run_name + "_param.txt", 'wt') as f: for key,value in sorted(params_lgb.items()): f.write(f'{key}:{value}\n') runner = Runner(run_name, ModelLGB, features, params_lgb, n_fold, name_prefix) runner.run_train_cv() # ֶश runner.run_predict_cv() # ਪ࿦ Submission.create_submission(run_name) # submit࡞੒ ͜ͷSVO@OBNFΛQSFpYͱͯ͠ɺϑΝΠϧ΍ϞσϧΛอଘͯ͘͠ΕΔɻ ྫ
    w ࢖༻ͨ͠ಛ௃ྔϦετ
    w ࢖༻ͨ͠ϋΠύʔύϥϝʔλ
    
    w GPMEຖͷϞσϧ
    w ਪ࿦݁Ռ
    w TVCNJUϑΝΠϧ
    w TIBQͷܭࢉ݁ՌΠϝʔδϑΝΠϧ
    
    
    ͳͲ
56. 56.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ SVOQZ
    Λ࣮ߦ͢Δ͜ͱͰɺֶशɾਪ࿦ɾ4VCNJUϑΝΠϧΛ࡞੒ # ಛ௃ྔͷࢦఆ features = [ "age_mis_val_median", "family__size",

    "cabin", "fare_mis_val_median" ] run_name = 'lgb_1102' # ࢖༻͢Δಛ௃ྔϦετͷอଘ with open(LOG_DIR_NAME + run_name + "_features.txt", 'wt') as f: for ele in features: f.write(ele+'\n') params_lgb = { 'boosting_type': 'gbdt', 'objective': 'binary', 'early_stopping_rounds': 20, 'verbose': 10, 'random_state': 99, 'num_round': 100 } # ࢖༻͢Δύϥϝʔλͷอଘ with open(LOG_DIR_NAME + run_name + "_param.txt", 'wt') as f: for key,value in sorted(params_lgb.items()): f.write(f'{key}:{value}\n') runner = Runner(run_name, ModelLGB, features, params_lgb, n_fold, name_prefix) runner.run_train_cv() # ֶश runner.run_predict_cv() # ਪ࿦ Submission.create_submission(run_name) # submit࡞੒ ੜ੒͞ΕΔϑΝΠϧྫ
57. 57.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ੜ੒͞ΕΔϑΝΠϧͷྫʢϑΥϧμ͸దٓ෼͚͍ͯ·͢ʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w

    MHC@@QSFEQLMʢUFTUσʔλͰͷਪ࿦݁Ռʣ
    w MHC@@@TVCNJTTJPODTWʢਪ࿦݁ՌΛLBHHMFʹఏग़Ͱ͖ΔDTWʹม׵ͨ͠΋ͷʣ
    w MHC@@@GFBUVSFTUYUʢࠓճͷֶशʹ࢖༻ͨ͠ಛ௃ྔϦετʣ
    w MHC@@@QBSBNUYUʢࠓճͷֶशʹ࢖༻ͨ͠ϋΠύʔύϥϝʔλ
    ʣ
    w MHC@@@TIBQQOHʢTIBQͰܭࢉͨ͠ՄࢹԽΠϝʔδʣ
    w HFOFSBMMPHʢܭࢉϩάϑΝΠϧʣ
    w SFTVMUMPHʢϞσϧͷείΞ͚͕ͩهࡌ͞ΕͨϩάϑΝΠϧʣ
58. 58.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ੜ੒͞ΕΔϑΝΠϧͷྫʢϑΥϧμ͸దٓ෼͚͍ͯ·͢ʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w

    MHC@@QSFEQLMʢUFTUσʔλͰͷਪ࿦݁Ռʣ
    w MHC@@@TVCNJTTJPODTWʢਪ࿦݁ՌΛLBHHMFʹఏग़Ͱ͖ΔDTWʹม׵ͨ͠΋ͷʣ
    w MHC@@@GFBUVSFTUYUʢࠓճͷֶशʹ࢖༻ͨ͠ಛ௃ྔϦετʣ
    w MHC@@@QBSBNUYUʢࠓճͷֶशʹ࢖༻ͨ͠ϋΠύʔύϥϝʔλ
    ʣ
    w MHC@@@TIBQQOHʢTIBQͰܭࢉͨ͠ՄࢹԽΠϝʔδʣ
    w HFOFSBMMPHʢܭࢉϩάϑΝΠϧʣ
    w SFTVMUMPHʢϞσϧͷείΞ͚͕ͩهࡌ͞ΕͨϩάϑΝΠϧʣ
59. 59.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ੜ੒͞ΕΔϑΝΠϧͷྫʢϑΥϧμ͸దٓ෼͚͍ͯ·͢ʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w

    MHC@@QSFEQLMʢUFTUσʔλͰͷਪ࿦݁Ռʣ
    w MHC@@@TVCNJTTJPODTWʢਪ࿦݁ՌΛLBHHMFʹఏग़Ͱ͖ΔDTWʹม׵ͨ͠΋ͷʣ
    w MHC@@@GFBUVSFTUYUʢࠓճͷֶशʹ࢖༻ͨ͠ಛ௃ྔϦετʣ
    w MHC@@@QBSBNUYUʢࠓճͷֶशʹ࢖༻ͨ͠ϋΠύʔύϥϝʔλ
    ʣ
    w MHC@@@TIBQQOHʢTIBQͰܭࢉͨ͠ՄࢹԽΠϝʔδʣ
    w HFOFSBMMPHʢܭࢉϩάϑΝΠϧʣ
    w SFTVMUMPHʢϞσϧͷείΞ͚͕ͩهࡌ͞ΕͨϩάϑΝΠϧʣ
60. 60.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ੜ੒͞ΕΔϑΝΠϧͷྫʢϑΥϧμ͸దٓ෼͚͍ͯ·͢ʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w

    MHC@@QSFEQLMʢUFTUσʔλͰͷਪ࿦݁Ռʣ
    w MHC@@@TVCNJTTJPODTWʢਪ࿦݁ՌΛLBHHMFʹఏग़Ͱ͖ΔDTWʹม׵ͨ͠΋ͷʣ
    w MHC@@@GFBUVSFTUYUʢࠓճͷֶशʹ࢖༻ͨ͠ಛ௃ྔϦετʣ
    w MHC@@@QBSBNUYUʢࠓճͷֶशʹ࢖༻ͨ͠ϋΠύʔύϥϝʔλ
    ʣ
    w MHC@@@TIBQQOHʢTIBQͰܭࢉͨ͠ՄࢹԽΠϝʔδʣ
    w HFOFSBMMPHʢܭࢉϩάϑΝΠϧʣ
    w SFTVMUMPHʢϞσϧͷείΞ͚͕ͩهࡌ͞ΕͨϩάϑΝΠϧʣ
61. 61.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ੜ੒͞ΕΔϑΝΠϧͷྫʢϑΥϧμ͸దٓ෼͚͍ͯ·͢ʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w

    MHC@@QSFEQLMʢUFTUσʔλͰͷਪ࿦݁Ռʣ
    w MHC@@@TVCNJTTJPODTWʢਪ࿦݁ՌΛLBHHMFʹఏग़Ͱ͖ΔDTWʹม׵ͨ͠΋ͷʣ
    w MHC@@@GFBUVSFTUYUʢࠓճͷֶशʹ࢖༻ͨ͠ಛ௃ྔϦετʣ
    w MHC@@@QBSBNUYUʢࠓճͷֶशʹ࢖༻ͨ͠ϋΠύʔύϥϝʔλ
    ʣ
    w MHC@@@TIBQQOHʢTIBQͰܭࢉͨ͠ՄࢹԽΠϝʔδʣ
    w HFOFSBMMPHʢܭࢉϩάϑΝΠϧʣ
    w SFTVMUMPHʢϞσϧͷείΞ͚͕ͩهࡌ͞ΕͨϩάϑΝΠϧʣ
62. 62.

    
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ ϚϚͷҰาΛࢧ͑Δ ੜ੒͞ΕΔϑΝΠϧͷྫʢϑΥϧμ͸దٓ෼͚͍ͯ·͢ʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w MHC@@GPMENPEFMʢGPMEͰ࡞੒͞ΕͨϞσϧʣ
    w

    MHC@@QSFEQLMʢUFTUσʔλͰͷਪ࿦݁Ռʣ
    w MHC@@@TVCNJTTJPODTWʢਪ࿦݁ՌΛLBHHMFʹఏग़Ͱ͖ΔDTWʹม׵ͨ͠΋ͷʣ
    w MHC@@@GFBUVSFTUYUʢࠓճͷֶशʹ࢖༻ͨ͠ಛ௃ྔϦετʣ
    w MHC@@@QBSBNUYUʢࠓճͷֶशʹ࢖༻ͨ͠ϋΠύʔύϥϝʔλ
    ʣ
    w MHC@@@TIBQQOHʢTIBQͰܭࢉͨ͠ՄࢹԽΠϝʔδʣ
    w HFOFSBMMPHʢܭࢉϩάϑΝΠϧʣ
    w SFTVMUMPHʢϞσϧͷείΞ͚͕ͩهࡌ͞ΕͨϩάϑΝΠϧʣ Կ͕خ͔͔ͬͨ͠
63. 63.

    ϚϚͷҰาΛࢧ͑Δ w ʮ͜ͷಛ௃ྔʯͱʮ͜ͷύϥϝʔλʯΛ࢖ֶͬͯशͤͨ͞Ϟσϧ ʹؔͯ͠ɺʮ֤λεΫʹཁͨ࣌ؒ͠ʯͱʮ֤GPMEʴ࠷ऴతͳεί ΞʯΛҙࣝ͠ͳͯ͘΋؅ཧͰ͖ΔΑ͏ʹɻ
    w TIBQͷܭࢉ݁Ռ΍GFBUVSF
    JNQPSUBODFΛग़ྗ͓ͯ͘͜͠ͱ Ͱɺ࣍ͷֶश࣌ͷצॴ͕௫ΊΔΑ͏ʹɻ 
    ֶशɾਪ࿦ύΠϓϥΠϯʹ͍ͭͯ

64. 64.

    ·ͱΊ ϚϚͷҰาΛࢧ͑Δ

65. 65.

    ·ͱΊ ϚϚͷҰาΛࢧ͑Δ w ಛ௃ྔ؅ཧ͍͍ͧʂ
     ͭͷεΫϦϓτϑΝΠϧʹಛ௃ྔੜ੒Λ·ͱΊΔ͜ͱͰɺಉ͡ܭࢉΛෳ਺ճ࣮ߦ͢Δ͜ͱ ΛճආͰ͖Δʂ
     ಛ௃ྔͷϝϞΛಉ࣌ʹੜ੒͢Δ͜ͱͰʮ͜ͷಛ௃ྔͳΜ͚ͩͬʁʯͱ಄Λ࢖͏ճ਺͕ݮ Δʂ
    

     ಛ௃ྔΛྻ͝ͱʹ؅ཧ͢Δ͜ͱͰऔΓճָ͕͠ʹͳͬͨʂʢ͕ɺಛ௃ྔ͕๲େͳ৔߹͸͋ Δఔ౓ͷ·ͱ·ΓͰ؅ཧͨ͠ํ͕ྑ͍͔΋ʣ
    w ύΠϓϥΠϯ͍͍ͧʂ
     ύΠϓϥΠϯΛߏங͢Δ͜ͱͰɺߴ଎ͳ1%$"Λ࣮ݱʂ
     ֶशʹ࢖༻ͨ͠ಛ௃ྔͱύϥϝʔλΛ؅ཧ͢Δ͜ͱͰɺ࠶ݱੑ΋୲อ͞Ε৺ཧత҆શੑ΋
66. 66.

    ϚϚͷҰาΛࢧ͑Δ ͝ਗ਼ௌ͋Γ͕ͱ͏͍͟͝·ͨ͠