[수업자료] ml2 오늘의 코드 -gridsearch, random, bagging

그리드 서치하는 이유: 최적의 파라미터 찾으려고. 

찾고나면 그리드 서치 지워버리고 파라미터만 적용하기 

 

ml13_gridSearchCV_iris

RandomForestClassifier()

param = [
    {'n_estimators': [100,200], 'max_depth' : [6,8,10],
     'n_jobs':[-1,1,2]},
    {'n_estimators': [100,500], 'min_samples_leaf':[3,5,7]}
]

#2. 모델
rf_model = RandomForestClassifier()
from sklearn.model_selection import GridSearchCV
model = GridSearchCV(rf_model, param, cv=kfold, verbose=1)

선생님이 임의로 설정한 수치의 파라미터. 

그리드서치하면 이 중에 최적의 값을 찾아서 출력해줌

Iris에 아래 파라미터 적용 후 돌

parameters = [
    {'n_estimators' : [100, 200], 'max_depth':[6, 8, 10, 12], 'n_jobs' : [-1, 2, 4]},  
    {'max_depth' : [6, 8, 10, 12], 'min_samples_split' : [2, 3, 5, 10]},
    {'n_estimators' : [100, 200], 'min_samples_leaf' : [3, 5, 7, 10]},
    {'min_samples_split' : [2, 3, 5, 10], 'n_jobs' : [-1, 2, 4]}, 
    {'n_estimators' : [100, 200],'n_jobs' : [-1, 2, 4]}

# 최적의 파라미터 :  {'max_depth': 6, 'n_estimators': 200, 'n_jobs': -1}

# 최적의 매개변수 :  RandomForestClassifier(max_depth=6, n_estimators=200, n_jobs=-1)

# best_score :  0.95

# model_score :  1.0

다른 파라미터랑 모델 수정

최적의 파라미터, 매개변수 찾기 

전

import numpy as np
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split, StratifiedKFold
from sklearn.model_selection import cross_val_score, cross_val_predict
from sklearn.datasets import load_iris
from sklearn.preprocessing import MinMaxScaler

import tensorflow as tf
tf.random.set_seed(77) # weight 난수값 조정

#1. 데이터
datasets = load_iris()
x = datasets['data']
y = datasets.target

x_train, x_test, y_train, y_test = train_test_split(
    x, y, train_size=0.8, shuffle=True, random_state=42
)

# kfold
n_splits = 5
random_state = 42
kfold = StratifiedKFold(n_splits=n_splits, shuffle=True, 
              random_state=random_state)

# Scaler 적용
scaler = MinMaxScaler()
scaler.fit(x_train)
x_train = scaler.transform(x_train)
x_test = scaler.transform(x_test)

param = [
    {'n_estimators' : [100, 200], 'max_depth':[6, 8, 10, 12], 'n_jobs' : [-1, 2, 4]},  
    {'max_depth' : [6, 8, 10, 12], 'min_samples_split' : [2, 3, 5, 10]},
    {'n_estimators' : [100, 200], 'min_samples_leaf' : [3, 5, 7, 10]},
    {'min_samples_split' : [2, 3, 5, 10], 'n_jobs' : [-1, 2, 4]}, 
    {'n_estimators' : [100, 200],'n_jobs' : [-1, 2, 4]}
]

#2. 모델
from sklearn.model_selection import GridSearchCV
rf_model = RandomForestClassifier()
model = GridSearchCV(rf_model, param, cv=kfold, verbose=1,
                     refit=True, n_jobs=-1) #refit 기본값은 False라 꼭 True로 해줘야함: 위에 입력한 파라미터를 찾아서 적용하는 기능

#3. 훈련
import time
start_time = time.time()
model.fit(x_train, y_train)
end_time= time.time() - start_time

print('최적의 파라미터 : ' , model.best_params_)
print('최적의 매개변수 : ' , model.best_estimator_)
print('best_score : ', model.best_score_)
print('model_score : ', model.score(x_test, y_test))
print('걸린 시간 : ', end_time, '초')

결과 값

최적의 파라미터 :  {'min_samples_leaf': 10, 'n_estimators': 200}
최적의 매개변수 :  RandomForestClassifier(min_samples_leaf=10, n_estimators=200)
best_score :  0.9583333333333334
model_score :  1.0
걸린 시간 :  10.046573162078857 초

결과 값 적용하는 코드 참고(최초 파라미터가 달라서 값이 다름 ㅜ )

import numpy as np
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split, StratifiedKFold
from sklearn.model_selection import cross_val_score, cross_val_predict
from sklearn.datasets import load_iris
from sklearn.preprocessing import MinMaxScaler

import tensorflow as tf
tf.random.set_seed(77) # weight 난수값 조정

#1. 데이터
datasets = load_iris()
x = datasets['data']
y = datasets.target

x_train, x_test, y_train, y_test = train_test_split(
    x, y, train_size=0.8, shuffle=True, random_state=42
)

# kfold
n_splits = 5
random_state = 42
kfold = StratifiedKFold(n_splits=n_splits, shuffle=True, 
              random_state=random_state)

# Scaler 적용
scaler = MinMaxScaler()
scaler.fit(x_train)
x_train = scaler.transform(x_train)
x_test = scaler.transform(x_test)


#2. 모델
from sklearn.model_selection import GridSearchCV
model = RandomForestClassifier(max_depth=6, n_estimators= 200, n_jobs= -1)
#refit 기본값은 False라 꼭 True로 해줘야함: 위에 입력한 파라미터를 찾아서 적용하는 기능

#3. 훈련
import time
start_time = time.time()
model.fit(x_train, y_train)
end_time= time.time() - start_time

print('걸린 시간 : ', end_time, '초')

# 최적의 파라미터 :  {'max_depth': 6, 'n_estimators': 200, 'n_jobs': -1}
# 최적의 매개변수 :  RandomForestClassifier(max_depth=6, n_estimators=200, n_jobs=-1)
# best_score :  0.95
# model_score :  1.0

# 그리드 서치 후
# 걸린 시간 :  0.2216627597808838 초
# cv pred acc :  0.9666666666666667


#4. 평가, 예측
score = cross_val_score(model, 
                        x_train, y_train, 
                        cv=kfold) # cv : cross validation
# print('cv acc : ', score)
y_predict = cross_val_predict(model,
                              x_test, y_test,
                              cv=kfold)
# print('cv pred : ', y_predict)
acc = accuracy_score(y_test, y_predict)
print('cv pred acc : ', acc)

# 결과 
# SVC() 결과 acc : 0.9777777777777777
# MinMaxScaler : 0.9777777777777777
# ==================================
# tree 결과 acc :  0.9555555555555556
# ==================================
# ensemble 결과 acc :  0.9555555555555556
# kfold 결과 acc :  1.0

ml13_gridSearchCV_california :

RandomForestRegressor, 그리드 돌리기

import numpy as np
from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split, StratifiedKFold, KFold
from sklearn.model_selection import cross_val_score, cross_val_predict
from sklearn.datasets import fetch_california_housing
from sklearn.preprocessing import MinMaxScaler

import tensorflow as tf
tf.random.set_seed(77) # weight 난수값 조정

#1. 데이터
datasets = fetch_california_housing()
x = datasets['data']
y = datasets.target

x_train, x_test, y_train, y_test = train_test_split(
    x, y, train_size=0.8, shuffle=True, random_state=42
)

# kfold
n_splits = 5
random_state = 42
kfold = KFold(n_splits=n_splits, shuffle=True, 
              random_state=random_state)

# Scaler 적용
scaler = MinMaxScaler()
scaler.fit(x_train)
x_train = scaler.transform(x_train)
x_test = scaler.transform(x_test)

param = [
    {'n_estimators' : [100, 200], 'max_depth':[6, 8, 10, 12], 'n_jobs' : [-1, 2, 4]},  
    {'max_depth' : [6, 8, 10, 12], 'min_samples_split' : [2, 3, 5, 10]},
    {'n_estimators' : [100, 200], 'min_samples_leaf' : [3, 5, 7, 10]},
    {'min_samples_split' : [2, 3, 5, 10], 'n_jobs' : [-1, 2, 4]}, 
    {'n_estimators' : [100, 200],'n_jobs' : [-1, 2, 4]}
]

#2. 모델
from sklearn.model_selection import GridSearchCV
rf_model = RandomForestRegressor()
model = GridSearchCV(rf_model, param, cv=kfold, verbose=1,
                     refit=True, n_jobs=-1) #refit 기본값은 False라 꼭 True로 해줘야함: 위에 입력한 파라미터를 찾아서 적용하는 기능

#3. 훈련
import time
start_time = time.time()
model.fit(x_train, y_train)
end_time= time.time() - start_time

print('최적의 파라미터 : ' , model.best_params_)
print('최적의 매개변수 : ' , model.best_estimator_)
print('best_score : ', model.best_score_)
print('model_score : ', model.score(x_test, y_test))
print('걸린 시간 : ', end_time, '초')

#4. 평가, 예측
score = cross_val_score(model, 
                        x_train, y_train, 
                        cv=kfold) # cv : cross validation
# print('cv acc : ', score)
y_predict = cross_val_predict(model,
                              x_test, y_test,
                              cv=kfold)
# print('cv pred : ', y_predict)
acc = accuracy_score(y_test, y_predict)
print('cv pred acc : ', acc)

그리드 결과

최적의 파라미터 :  {'n_estimators': 200, 'n_jobs': 4}
최적의 매개변수 :  RandomForestRegressor(n_estimators=200, n_jobs=4)
best_score :  0.8057484561327181
model_score :  0.8088843983852754
걸린 시간 :  466.20924973487854 초

너무 오래걸려서 랜덤그리드서치로 가야함 ,, ,!!! 

 

랜덤서치코드

#2. 모델
from sklearn.model_selection import RandomizedSearchCV
rf_model = RandomForestRegressor()
model = RandomizedSearchCV(rf_model, param, cv=kfold, verbose=1,
                     refit=True, n_jobs=-1) 
                     #refit 기본값은 False라 꼭 True로 해줘야함: 위에 입력한 파라미터를 찾아서 적용하는 기능

랜덤서치값

최적의 파라미터 :  {'n_estimators': 200, 'n_jobs': -1}
최적의 매개변수 :  RandomForestRegressor(n_estimators=200, n_jobs=-1)
best_score :  0.8057475283936508
model_score :  0.8070839473608384
걸린 시간 :  698.0324904918671 초

xgboost 모델의 parameters

ml15_gridSearchCV_xgb_iris - Jupyter Notebook

그리드 적용 result : ml15_gridSearchCV_xgb_iris_result - Jupyter Notebook

LGBM 모델의 parameters -실습 안 했지만 팀플에 쓰기 

catboost 모델의 parameters-실습 안 했지만 팀플에 쓰기 

ml16_gridSearchCV_cat_iris - Jupyter Notebook

 

 

ml16_bagging_iris

원래 결과값이 좋은 데이터.

bagging : 성능 좋음 

import numpy as np
from sklearn.model_selection import train_test_split, StratifiedKFold
from sklearn.model_selection import GridSearchCV
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import BaggingClassifier
from sklearn.preprocessing import MinMaxScaler
from sklearn.datasets import load_iris
import time

#1. 데이터
datasets = load_iris()
x = datasets.data
y = datasets.target

x_train,x_test, y_train, y_test = train_test_split(x,y,train_size=0.8, random_state=42, shuffle=True)

# 스케일러
scaler = MinMaxScaler()
x_train = scaler.fit_transform(x_train)
x_test = scaler.transform(x_test)

#2. 모델(Bagging)
bagging = BaggingClassifier(DecisionTreeClassifier(), 
                            n_estimators=100,
                            n_jobs=-1,
                            random_state=42)

#3. 훈련
start_time = time.time()
bagging.fit(x_train,y_train)
end_time = time.time() - start_time

#4. 평가,예측
result = bagging.score(x_test,y_test)

print('걸린시간: ', end_time, '초')
print('bagging 결과 : ', result)

걸린시간:  1.556366205215454 초
bagging 결과 :  1.0

ml16_bagging_cali

 

import numpy as np
from sklearn.model_selection import train_test_split, KFold
from sklearn.model_selection import GridSearchCV
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import BaggingRegressor
from sklearn.preprocessing import MinMaxScaler
from sklearn.datasets import fetch_california_housing
import time

#1. 데이터
datasets = fetch_california_housing()
x = datasets.data
y = datasets.target

x_train,x_test, y_train, y_test = train_test_split(x,y,train_size=0.8, random_state=42, shuffle=True)

# 스케일러
scaler = MinMaxScaler()
x_train = scaler.fit_transform(x_train)
x_test = scaler.transform(x_test)

#2. 모델(Bagging)
bagging = BaggingRegressor(DecisionTreeRegressor(), 
                            n_estimators=100,
                            n_jobs=-1,
                            random_state=42)

#3. 훈련
start_time = time.time()
bagging.fit(x_train,y_train)
end_time = time.time() - start_time

#4. 평가,예측
result = bagging.score(x_test,y_test)

print('걸린시간: ', end_time, '초')
print('bagging 결과 : ', result)

# 걸린시간:  3.731471300125122 초
# bagging 결과 :  0.8042348744721923

배깅 + kfold, grid

import numpy as np
from sklearn.model_selection import train_test_split, KFold
from sklearn.model_selection import GridSearchCV
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import BaggingRegressor
from sklearn.preprocessing import MinMaxScaler
from sklearn.datasets import fetch_california_housing
import time

#1. 데이터
datasets = fetch_california_housing()
x = datasets.data
y = datasets.target

x_train,x_test, y_train, y_test = train_test_split(x,y,train_size=0.8, random_state=42, shuffle=True)


# 스케일러
scaler = MinMaxScaler()
x_train = scaler.fit_transform(x_train)
x_test = scaler.transform(x_test)

#KFold
n_split = 5
kfold = KFold(n_splits=n_split, shuffle=True, random_state=42)

#parameter
param ={
    'n_estimators': [100],
    'random_state':[42,62,72],
    'max_features':[3,4,7]
}

#2. 모델(Bagging)
bagging = BaggingRegressor(DecisionTreeRegressor(), 
                            n_estimators=100,
                            n_jobs=-1,
                            random_state=42)
model = GridSearchCV(bagging, param, cv=kfold, refit=True, n_jobs=-1)

#3. 훈련
start_time = time.time()
model.fit(x_train,y_train)
end_time = time.time() - start_time

#4. 평가,예측
result = model.score(x_test,y_test)

print('최적의 매개변수 : ', model.best_estimator_)
print('최적의 파라미터 : ', model.best_params_)
print('걸린시간: ', end_time, '초')
print('bagging 결과 : ', result)

#배깅만 했을 때
# 걸린시간:  3.731471300125122 초
# bagging 결과 :  0.8042348744721923

#kfold, grid 적용결과
# 최적의 매개변수 :  BaggingRegressor(estimator=DecisionTreeRegressor(), max_features=7,
#                  n_estimators=100, n_jobs=-1, random_state=42)
# 최적의 파라미터 :  {'max_features': 7, 'n_estimators': 100, 'random_state': 42}
# 걸린시간:  49.88687562942505 초
# bagging 결과 :  0.822087951242456

파라미터 & 배깅 & kfold, grid 모두 넣으면 값이 더 좋아짐 ! 

 

voting : boost 여러 개의 다른 모델을 조합하여 더 강력하고 정확한 예측 모델을 만드는 방법

hard와 soft 차이 있음

 

iris voting classifier 

import numpy as np 
from sklearn.model_selection import train_test_split, StratifiedKFold
from sklearn.model_selection import GridSearchCV
from sklearn.preprocessing import MinMaxScaler
from sklearn.datasets import load_iris
from sklearn.ensemble import VotingClassifier
from sklearn.metrics import accuracy_score
from xgboost import XGBClassifier
from lightgbm import LGBMClassifier
from catboost import CatBoostClassifier

#1. 데이터 
datasets = load_iris()
x = datasets.data
y = datasets.target

x_train, x_test, y_train, y_test = train_test_split(x,y,test_size=0.2, shuffle=True, random_state=42)

# 스케일러
scaler = MinMaxScaler()
x_train = scaler.fit_transform(x_train)
x_test = scaler.transform(x_test)

#2. 모델(voting)
xgb = XGBClassifier()
lgbm = LGBMClassifier()
cat = CatBoostClassifier()

model = VotingClassifier(estimators=[('xgb', xgb),('lgbm', lgbm),('cat', cat)],# 모델이랑 이름 같이 넣어줘야함
                         voting = 'hard',
                         n_jobs=-1) 

#3. 훈련
model.fit(x_train,y_train)

# 4. 평가 예측
# y_predict = model.predict(x_test)
# score = accuracy_score(y_test, y_predict)
# print('voting 결과: ', score)

classifiers = [cat,xgb,lgbm,]
for model in classifiers:
    model.fit(x_train,y_train)
    y_predict=model.predict(x_test)
    score = accuracy_score(y_test, y_predict)
    class_name = model.__class__.__name__
    print('{0} 정확도: {1: .4f}'.format(class_name, score))
    
# CatBoostClassifier 정확도:  1.0000
# XGBClassifier 정확도:  1.0000
# LGBMClassifier 정확도:  1.0000

california voting regressor 

import numpy as np 
from sklearn.model_selection import train_test_split, KFold
from sklearn.model_selection import GridSearchCV
from sklearn.preprocessing import MinMaxScaler
from sklearn.datasets import fetch_california_housing
from sklearn.ensemble import VotingRegressor
from sklearn.metrics import r2_score
from xgboost import XGBRegressor
from lightgbm import LGBMRegressor
from catboost import CatBoostRegressor

#1. 데이터 
datasets = fetch_california_housing()
x = datasets.data
y = datasets.target

x_train, x_test, y_train, y_test = train_test_split(x,y,test_size=0.2, shuffle=True, random_state=42)

# 스케일러
scaler = MinMaxScaler()
x_train = scaler.fit_transform(x_train)
x_test = scaler.transform(x_test)

#2. 모델(voting)-모델에 그리드서치한 값을 각 클래스에 추가해서 돌리기 
xgb = XGBRegressor()
lgbm = LGBMRegressor()
cat = CatBoostRegressor()

model = VotingRegressor(estimators=[('xgb', xgb),('lgbm', lgbm),('cat', cat)], # 모델이랑 이름 같이 넣어줘야함
                         n_jobs=-1) #regressor에선 괄호안에 voting 없음 , 하드 소프트는: 분류 모델에서만 사용. 

#3. 훈련
model.fit(x_train,y_train)

# 4. 평가 예측
# y_predict = model.predict(x_test)
# score = accuracy_score(y_test, y_predict)
# print('voting 결과: ', score)

regressors = [cat,xgb,lgbm,]
for model in regressors:
    model.fit(x_train,y_train)
    y_predict=model.predict(x_test)
    score = r2_score(y_test, y_predict)
    class_name = model.__class__.__name__
    print('{0} 정확도: {1: .4f}'.format(class_name, score))

#회귀는 hard soft 구분이 없음 
# CatBoostRegressor 정확도:  0.8492
# XGBRegressor 정확도:  0.8287
# LGBMRegressor 정확도:  0.8365

아웃라이어 - iqr 중심으로 보기, 손코딩, 시각화 

import numpy as np

oliers = np.array([-50,-10,2,3,4,5,6,7,8,9,10,11,12,50])

def outliers (data_out):
    quartile_1, q2, quartile_3 = np.percentile(data_out,
                                              [25,50,75])       #등분하는거
    print('1사분위 : ', quartile_1)
    print('2사분위 : ', q2)
    print('3사분위 : ', quartile_3)
    
    iqr = quartile_3 - quartile_1
    print('IQR : ',iqr)
    
    lower_bound = quartile_1 - (iqr*1.5)    #낮은 이상치
    upper_bound = quartile_3 + (iqr*1.5)    #높은 이상치 
    print('lower_bound : ', lower_bound)
    print('upper_bound : ', upper_bound)
    return np.where((data_out>upper_bound) |
                    (data_out<lower_bound))
    
outliers_loc = outliers(oliers)
print('이상치의 위치 : ', outliers_loc)
***********
1사분위 :  3.25
2사분위 :  6.5
3사분위 :  9.75
IQR :  6.5
lower_bound :  -6.5
upper_bound :  19.5
이상치의 위치 :  (array([ 0,  1, 13], dtype=int64),)

컬럼별로 넣어서 이상치 확인 후 제거할 부분 제거 ;

#시각화 
import matplotlib.pyplot as plt
plt.boxplot(oliers)
plt.show()

=> iqr 벗어나는 부분 삭제해도 됨 ; 

 

ml18_outliers.EllipticEnvelope.py

import numpy as np

oliers = np.array([-50,-10,2,3,4,5,6,7,8,9,10,11,12,50])
print(oliers.shape)     #(14,)
oliers = oliers.reshape(-1,1)
print(oliers.shape)     #(14, 1)

from sklearn.covariance import EllipticEnvelope
outliers = EllipticEnvelope(contamination =.1) #이상치 몇 퍼 지정할지 정하기

outliers.fit(oliers)
result = outliers.predict(oliers)

print(result)
print(result.shape) #(14,)
#[-1  1  1  1  1  1  1  1  1  1  1  1  1 -1]v

결측치 확인

import numpy as np
import pandas as pd

data = pd.DataFrame([[2, np.nan, 6,8,10],
                    [2,4,np.nan, 8,np.nan,],
                    [2,4,6,8,10],
                    [np.nan,4, np.nan, 8, np.nan]])
# print(data)
# print(data.shape)
#      0    1    2  3     4
# 0  2.0  NaN  6.0  8  10.0
# 1  2.0  4.0  NaN  8   NaN
# 2  2.0  4.0  6.0  8  10.0
# 3  NaN  4.0  NaN  8   NaN
# (4, 5)

data = data.transpose()
data.columns = ['x1','x2','x3', 'x4']

# print(data)
# print(data.shape)

#      x1   x2    x3   x4
# 0   2.0  2.0   2.0  NaN
# 1   NaN  4.0   4.0  4.0
# 2   6.0  NaN   6.0  NaN
# 3   8.0  8.0   8.0  8.0
# 4  10.0  NaN  10.0  NaN
# (5, 4)

#결측치 확인
print(data.isnull())
print(data.isnull().sum())
print(data.info())

# 1. 결측치 삭제 
print(data.dropna(axis=0)) #난 값이 들어있는 컬럼이 다 없어짐 
print(data.shape)
'''
# print(data.dropna()) 
#     x1   x2   x3   x4
# 3  8.0  8.0  8.0  8.0
# (5, 4)

# axis=0이면 컬럼 다 살아있는데 행이 사라짐 
    x1   x2   x3   x4
3  8.0  8.0  8.0  8.0
(5, 4)

# axis=1이면 컬럼 열이 다 사라짐 
     x3
0   2.0
1   4.0
2   6.0
3   8.0
4  10.0
(5, 4)
'''
'''
#2. 특정값으로 대체 
means = data.mean()     # 평균 
median = data.median()  #중간 값 

data2 = data.fillna(means)
print(data2)
     x1        x2    x3   x4
0   2.0  2.000000   2.0  6.0
1   6.5  4.000000   4.0  4.0
2   6.0  4.666667   6.0  6.0
3   8.0  8.000000   8.0  8.0
4  10.0  4.666667  10.0  6.0

data3= data.fillna(median)
print(data3)
     x1   x2    x3   x4
0   2.0  2.0   2.0  6.0
1   7.0  4.0   4.0  4.0
2   6.0  4.0   6.0  6.0
3   8.0  8.0   8.0  8.0
4  10.0  4.0  10.0  6.0
'''

sklearn으로 결측지 값 정리 

import numpy as np
import pandas as pd

data = pd.DataFrame([[2, np.nan, 6,8,10],
                    [2,4,np.nan, 8,np.nan,],
                    [2,4,6,8,10],
                    [np.nan,4, np.nan, 8, np.nan]])

data = data.transpose()
data.columns = ['x1','x2','x3', 'x4']

# from sklearn.impute import IterativeImputer
# from sklearn.experimental import enable_iterative_imputer
from sklearn.impute import SimpleImputer

# imputer = SimpleImputer()   #평균값으로 대체(default)
# imputer = SimpleImputer(strategy='mean') #평균값 
# imputer = SimpleImputer(strategy='median') # 중간값 
# imputer = SimpleImputer(strategy='most_frequent')   #가장 많이 사용된 값 
imputer = SimpleImputer(strategy='constant', fill_value=777) #상수 입력값, default = 0 
imputer.fit(data)
data2 = imputer.transform(data)
print(data2)
'''
imputer = SimpleImputer()
[[ 2.          2.          2.          6.        ]
 [ 6.5         4.          4.          4.        ]
 [ 6.          4.66666667  6.          6.        ]
 [ 8.          8.          8.          8.        ]
 [10.          4.66666667 10.          6.        ]]
 
 imputer = SimpleImputer(strategy='median') # 중간값
 [[ 2.  2.  2.  6.]
 [ 7.  4.  4.  4.]
 [ 6.  4.  6.  6.]
 [ 8.  8.  8.  8.]
 [10.  4. 10.  6.]]
 
 imputer = SimpleImputer(strategy='most_frequent')
 [[ 2.  2.  2.  4.]
 [ 2.  4.  4.  4.]
 [ 6.  2.  6.  4.]
 [ 8.  8.  8.  8.]
 [10.  2. 10.  4.]]
 
 imputer = SimpleImputer(strategy='constant', fill_value=777) 
 [[  2.   2.   2. 777.]
 [777.   4.   4.   4.]
 [  6. 777.   6. 777.]
 [  8.   8.   8.   8.]
 [ 10. 777.  10. 777.]]
 '''

데이터 확인 후 숫자 너무 큰건 줄여주기 

#log 변환
import numpy as np
import pandas as pd
from sklearn.metrics import r2_score
from sklearn.datasets import fetch_california_housing
from sklearn.model_selection import train_test_split, KFold
from sklearn.model_selection import cross_val_score, cross_val_predict
from sklearn.preprocessing import MinMaxScaler, StandardScaler
from sklearn.preprocessing import MaxAbsScaler, RobustScaler


#1. 데이터
datasets = fetch_california_housing()
x = datasets['data']
y = datasets.target

df = pd.DataFrame(x, columns=[datasets.feature_names])
print(df)
print(df.head())
print(df['Population']) #다른데이터에 비해 숫자가 너무 큼 다른건 다 2자리 이하인데 이건 4자리

df['Population'] = np.log1p(df['Population'])
print(df['Population'].head())

  Population
0   5.777652
1   7.784057
2   6.208590
3   6.326149
4   6.338594