사이킷런 (scikit-learn,sklearn, diabetes)

scikit-learn(sklearn)

by DeepLearning Engineer 2021. 7. 12. 10:16

SMALL

사이킷런(scikit-learn / sklearn) - diabetes datasets¶

이번 포스팅에서는 앞에사용했던 기법들을 이용해서 diabetes 데이터셋을 분석해 보겠습니다.

필요모듈 import¶

In [1]:

import numpy as np
import pandas as pd
from sklearn.datasets import load_diabetes
from sklearn.model_selection import train_test_split
from sklearn.metrics import r2_score,mean_squared_error
from sklearn.preprocessing import StandardScaler,MinMaxScaler
from sklearn.utils import all_estimators
import warnings
warnings.filterwarnings('ignore')

In [2]:

datasets = load_diabetes()
x = datasets.data
y = datasets.target.reshape(-1,1)
print(x.shape,y.shape)

(442, 10) (442, 1)

이전과 마찬가지로 데이터셋을 로드해준 후에 타겟 데이터와 feature데이터셋을 각각 x,y로 지정해 주었습니다. diabetes 데이터셋은 442개의 행으로 이루어져 있네요.

In [3]:

print(datasets.feature_names)
print(datasets.DESCR)

['age', 'sex', 'bmi', 'bp', 's1', 's2', 's3', 's4', 's5', 's6']
.. _diabetes_dataset:

Diabetes dataset
----------------

Ten baseline variables, age, sex, body mass index, average blood
pressure, and six blood serum measurements were obtained for each of n =
442 diabetes patients, as well as the response of interest, a
quantitative measure of disease progression one year after baseline.

**Data Set Characteristics:**

  :Number of Instances: 442

  :Number of Attributes: First 10 columns are numeric predictive values

  :Target: Column 11 is a quantitative measure of disease progression one year after baseline

  :Attribute Information:
      - age     age in years
      - sex
      - bmi     body mass index
      - bp      average blood pressure
      - s1      tc, T-Cells (a type of white blood cells)
      - s2      ldl, low-density lipoproteins
      - s3      hdl, high-density lipoproteins
      - s4      tch, thyroid stimulating hormone
      - s5      ltg, lamotrigine
      - s6      glu, blood sugar level

Note: Each of these 10 feature variables have been mean centered and scaled by the standard deviation times `n_samples` (i.e. the sum of squares of each column totals 1).

Source URL:
https://www4.stat.ncsu.edu/~boos/var.select/diabetes.html

For more information see:
Bradley Efron, Trevor Hastie, Iain Johnstone and Robert Tibshirani (2004) "Least Angle Regression," Annals of Statistics (with discussion), 407-499.
(https://web.stanford.edu/~hastie/Papers/LARS/LeastAngle_2002.pdf)

In [4]:

columns = datasets.feature_names
columns.append("Target(Diabetes dataset)")

data = np.concatenate([x,y],axis=1)
dataframe = pd.DataFrame(data,columns = columns)
dataframe

Out[4]:

	age	sex	bmi	bp	s1	s2	s3	s4	s5	s6	Target(Diabetes dataset)
0	0.038076	0.050680	0.061696	0.021872	-0.044223	-0.034821	-0.043401	-0.002592	0.019908	-0.017646	151.0
1	-0.001882	-0.044642	-0.051474	-0.026328	-0.008449	-0.019163	0.074412	-0.039493	-0.068330	-0.092204	75.0
2	0.085299	0.050680	0.044451	-0.005671	-0.045599	-0.034194	-0.032356	-0.002592	0.002864	-0.025930	141.0
3	-0.089063	-0.044642	-0.011595	-0.036656	0.012191	0.024991	-0.036038	0.034309	0.022692	-0.009362	206.0
4	0.005383	-0.044642	-0.036385	0.021872	0.003935	0.015596	0.008142	-0.002592	-0.031991	-0.046641	135.0
...	...	...	...	...	...	...	...	...	...	...	...
437	0.041708	0.050680	0.019662	0.059744	-0.005697	-0.002566	-0.028674	-0.002592	0.031193	0.007207	178.0
438	-0.005515	0.050680	-0.015906	-0.067642	0.049341	0.079165	-0.028674	0.034309	-0.018118	0.044485	104.0
439	0.041708	0.050680	-0.015906	0.017282	-0.037344	-0.013840	-0.024993	-0.011080	-0.046879	0.015491	132.0
440	-0.045472	-0.044642	0.039062	0.001215	0.016318	0.015283	-0.028674	0.026560	0.044528	-0.025930	220.0
441	-0.045472	-0.044642	-0.073030	-0.081414	0.083740	0.027809	0.173816	-0.039493	-0.004220	0.003064	57.0

442 rows × 11 columns

데이터셋을 보면 우리가 예측해야할 값이 클래스변수(분류를 해야하는)가 아닌 수치인 것으로 보아 회귀문제네요.

마찬가지로 데이터셋을 위의 모양에서 다시 잘라주겠습니다.

In [5]:

datasets = dataframe.values

x = datasets[:,:-1]
y = datasets[:,-1]
print(x.shape,y.shape)

(442, 10) (442,)

train_test_split 함수를 통해서 해당 데이터셋을 train/test으로 나누겠습니다.¶

In [6]:

x_train,x_test,y_train,y_test = train_test_split(x,y,test_size=0.2)

print(x_train.shape,x_test.shape)
print(y_train.shape,y_test.shape)

(353, 10) (89, 10)
(353,) (89,)

이전과 마찬가지로 all_estimator를 이용해서 대략적으로 어느정도의 값이 나오는지 확인해 봅니다.¶

또한 이번에는 standard scaler를 포함시켜 minmax 와 standard scaler중에 누가 더 좋은 성능을 내는지 확인해 봅시다.¶

In [10]:

all_Algorithm = all_estimators(type_filter = 'regressor')
scaler_list = [StandardScaler(),MinMaxScaler()]
best_r2=[]
for scaler in scaler_list:
    scaler.fit(x_train)
    x_train = scaler.transform(x_train)
    x_test = scaler.transform(x_test)
    for (name,algorithm) in all_Algorithm:
        try:
            model = algorithm()
            model.fit(x_train,y_train)
            y_pred = model.predict(x_test)
            r2 = r2_score(y_test,y_pred)
            best_r2.append(r2)
            print(name , "\'s R2 : ", r2)
            print("-----------------------------------------------")
            print(r2,max(best_r2))
            if max(best_r2)<=r2:
                final_r2 = r2
                final_r2_model = name
                final_r2_scaler = scaler
        except:
            continue
        
print("Best R2 & model & scaler: ",final_r2," & ",final_r2_model, "&",scaler)

ARDRegression 's R2 :  0.3750119550784249
-----------------------------------------------
0.3750119550784249 0.3750119550784249
AdaBoostRegressor 's R2 :  0.35328796754369163
-----------------------------------------------
0.35328796754369163 0.3750119550784249
BaggingRegressor 's R2 :  0.31602822849489987
-----------------------------------------------
0.31602822849489987 0.3750119550784249
BayesianRidge 's R2 :  0.37353709794963996
-----------------------------------------------
0.37353709794963996 0.3750119550784249
CCA 's R2 :  0.3645458498695413
-----------------------------------------------
0.3645458498695413 0.3750119550784249
DecisionTreeRegressor 's R2 :  -0.5894581868323021
-----------------------------------------------
-0.5894581868323021 0.3750119550784249
DummyRegressor 's R2 :  -0.00123645803999195
-----------------------------------------------
-0.00123645803999195 0.3750119550784249
ElasticNet 's R2 :  0.3790340028177188
-----------------------------------------------
0.3790340028177188 0.3790340028177188
ElasticNetCV 's R2 :  0.3754605025485025
-----------------------------------------------
0.3754605025485025 0.3790340028177188
ExtraTreeRegressor 's R2 :  -0.3050881139739763
-----------------------------------------------
-0.3050881139739763 0.3790340028177188
ExtraTreesRegressor 's R2 :  0.3497806996920655
-----------------------------------------------
0.3497806996920655 0.3790340028177188
GammaRegressor 's R2 :  0.3701174806186931
-----------------------------------------------
0.3701174806186931 0.3790340028177188
GaussianProcessRegressor 's R2 :  -1.069012594401333
-----------------------------------------------
-1.069012594401333 0.3790340028177188
GradientBoostingRegressor 's R2 :  0.33250985042993264
-----------------------------------------------
0.33250985042993264 0.3790340028177188
HistGradientBoostingRegressor 's R2 :  0.2751147773065651
-----------------------------------------------
0.2751147773065651 0.3790340028177188
HuberRegressor 's R2 :  0.34174750477637783
-----------------------------------------------
0.34174750477637783 0.3790340028177188
KNeighborsRegressor 's R2 :  0.25446926061787567
-----------------------------------------------
0.25446926061787567 0.3790340028177188
KernelRidge 's R2 :  -4.165048939009507
-----------------------------------------------
-4.165048939009507 0.3790340028177188
Lars 's R2 :  0.34837209873152053
-----------------------------------------------
0.34837209873152053 0.3790340028177188
LarsCV 's R2 :  0.40257159246468
-----------------------------------------------
0.40257159246468 0.40257159246468
Lasso 's R2 :  0.38595024820903034
-----------------------------------------------
0.38595024820903034 0.40257159246468
LassoCV 's R2 :  0.35338208122936043
-----------------------------------------------
0.35338208122936043 0.40257159246468
LassoLars 's R2 :  0.35382890509034337
-----------------------------------------------
0.35382890509034337 0.40257159246468
LassoLarsCV 's R2 :  0.3531078796996593
-----------------------------------------------
0.3531078796996593 0.40257159246468
LassoLarsIC 's R2 :  0.39914059916427813
-----------------------------------------------
0.39914059916427813 0.40257159246468
LinearRegression 's R2 :  0.34837209873152
-----------------------------------------------
0.34837209873152 0.40257159246468
LinearSVR 's R2 :  0.23760072727231019
-----------------------------------------------
0.23760072727231019 0.40257159246468
MLPRegressor 's R2 :  -1.25918081681785
-----------------------------------------------
-1.25918081681785 0.40257159246468
NuSVR 's R2 :  0.14418144713660686
-----------------------------------------------
0.14418144713660686 0.40257159246468
OrthogonalMatchingPursuit 's R2 :  0.30876981915826296
-----------------------------------------------
0.30876981915826296 0.40257159246468
OrthogonalMatchingPursuitCV 's R2 :  0.34286930368891755
-----------------------------------------------
0.34286930368891755 0.40257159246468
PLSCanonical 's R2 :  -2.0847428167722253
-----------------------------------------------
-2.0847428167722253 0.40257159246468
PLSRegression 's R2 :  0.33865842534282786
-----------------------------------------------
0.33865842534282786 0.40257159246468
PassiveAggressiveRegressor 's R2 :  0.30424743439568547
-----------------------------------------------
0.30424743439568547 0.40257159246468
PoissonRegressor 's R2 :  0.39452724195999944
-----------------------------------------------
0.39452724195999944 0.40257159246468
RANSACRegressor 's R2 :  0.07825356015037255
-----------------------------------------------
0.07825356015037255 0.40257159246468
RandomForestRegressor 's R2 :  0.3917103268004255
-----------------------------------------------
0.3917103268004255 0.40257159246468
Ridge 's R2 :  0.3564366035510431
-----------------------------------------------
0.3564366035510431 0.40257159246468
RidgeCV 's R2 :  0.3494465130272413
-----------------------------------------------
0.3494465130272413 0.40257159246468
SGDRegressor 's R2 :  0.37452975226115104
-----------------------------------------------
0.37452975226115104 0.40257159246468
SVR 's R2 :  0.13622570343080198
-----------------------------------------------
0.13622570343080198 0.40257159246468
TheilSenRegressor 's R2 :  0.36489787952814867
-----------------------------------------------
0.36489787952814867 0.40257159246468
TransformedTargetRegressor 's R2 :  0.34837209873152
-----------------------------------------------
0.34837209873152 0.40257159246468
TweedieRegressor 's R2 :  0.36243293782488917
-----------------------------------------------
0.36243293782488917 0.40257159246468
ARDRegression 's R2 :  0.3750102764327631
-----------------------------------------------
0.3750102764327631 0.40257159246468
AdaBoostRegressor 's R2 :  0.37264064070142944
-----------------------------------------------
0.37264064070142944 0.40257159246468
BaggingRegressor 's R2 :  0.31819177735841186
-----------------------------------------------
0.31819177735841186 0.40257159246468
BayesianRidge 's R2 :  0.3752167808635509
-----------------------------------------------
0.3752167808635509 0.40257159246468
CCA 's R2 :  0.3645458498695415
-----------------------------------------------
0.3645458498695415 0.40257159246468
DecisionTreeRegressor 's R2 :  -0.5959577953403712
-----------------------------------------------
-0.5959577953403712 0.40257159246468
DummyRegressor 's R2 :  -0.00123645803999195
-----------------------------------------------
-0.00123645803999195 0.40257159246468
ElasticNet 's R2 :  0.12758535659494175
-----------------------------------------------
0.12758535659494175 0.40257159246468
ElasticNetCV 's R2 :  0.3871581150993674
-----------------------------------------------
0.3871581150993674 0.40257159246468
ExtraTreeRegressor 's R2 :  -0.3061839524050296
-----------------------------------------------
-0.3061839524050296 0.40257159246468
ExtraTreesRegressor 's R2 :  0.388669846985901
-----------------------------------------------
0.388669846985901 0.40257159246468
GammaRegressor 's R2 :  0.0808616210477957
-----------------------------------------------
0.0808616210477957 0.40257159246468
GaussianProcessRegressor 's R2 :  -18.53318219435217
-----------------------------------------------
-18.53318219435217 0.40257159246468
GradientBoostingRegressor 's R2 :  0.32579797431094626
-----------------------------------------------
0.32579797431094626 0.40257159246468
HistGradientBoostingRegressor 's R2 :  0.27523810100653423
-----------------------------------------------
0.27523810100653423 0.40257159246468
HuberRegressor 's R2 :  0.34367339121978224
-----------------------------------------------
0.34367339121978224 0.40257159246468
KNeighborsRegressor 's R2 :  0.24594913899940074
-----------------------------------------------
0.24594913899940074 0.40257159246468
KernelRidge 's R2 :  0.3628155857813502
-----------------------------------------------
0.3628155857813502 0.40257159246468
Lars 's R2 :  0.3483720987315203
-----------------------------------------------
0.3483720987315203 0.40257159246468
LarsCV 's R2 :  0.40257159246468
-----------------------------------------------
0.40257159246468 0.40257159246468
Lasso 's R2 :  0.416032632027838
-----------------------------------------------
0.416032632027838 0.416032632027838
LassoCV 's R2 :  0.3540874239008369
-----------------------------------------------
0.3540874239008369 0.416032632027838
LassoLars 's R2 :  0.3538289050903437
-----------------------------------------------
0.3538289050903437 0.416032632027838
LassoLarsCV 's R2 :  0.3531078796996586
-----------------------------------------------
0.3531078796996586 0.416032632027838
LassoLarsIC 's R2 :  0.39914059916427813
-----------------------------------------------
0.39914059916427813 0.416032632027838
LinearRegression 's R2 :  0.3483720987315203
-----------------------------------------------
0.3483720987315203 0.416032632027838
LinearSVR 's R2 :  0.18218041356580628
-----------------------------------------------
0.18218041356580628 0.416032632027838
MLPRegressor 's R2 :  -0.7391046842747242
-----------------------------------------------
-0.7391046842747242 0.416032632027838
NuSVR 's R2 :  0.1330302415587895
-----------------------------------------------
0.1330302415587895 0.416032632027838
OrthogonalMatchingPursuit 's R2 :  0.30876981915826285
-----------------------------------------------
0.30876981915826285 0.416032632027838
OrthogonalMatchingPursuitCV 's R2 :  0.3428693036889173
-----------------------------------------------
0.3428693036889173 0.416032632027838
PLSCanonical 's R2 :  -2.0847428167722244
-----------------------------------------------
-2.0847428167722244 0.416032632027838
PLSRegression 's R2 :  0.33865842534282786
-----------------------------------------------
0.33865842534282786 0.416032632027838
PassiveAggressiveRegressor 's R2 :  0.33870828293724464
-----------------------------------------------
0.33870828293724464 0.416032632027838
PoissonRegressor 's R2 :  0.3892955288595463
-----------------------------------------------
0.3892955288595463 0.416032632027838
RANSACRegressor 's R2 :  0.011896314581101852
-----------------------------------------------
0.011896314581101852 0.416032632027838
RadiusNeighborsRegressor 's R2 :  0.1656451755381757
-----------------------------------------------
0.1656451755381757 0.416032632027838
RandomForestRegressor 's R2 :  0.35712332450451356
-----------------------------------------------
0.35712332450451356 0.416032632027838
Ridge 's R2 :  0.3795714326643239
-----------------------------------------------
0.3795714326643239 0.416032632027838
RidgeCV 's R2 :  0.36502947676996134
-----------------------------------------------
0.36502947676996134 0.416032632027838
SGDRegressor 's R2 :  0.3790051512357022
-----------------------------------------------
0.3790051512357022 0.416032632027838
SVR 's R2 :  0.13059068934603013
-----------------------------------------------
0.13059068934603013 0.416032632027838
TheilSenRegressor 's R2 :  0.3668935873138036
-----------------------------------------------
0.3668935873138036 0.416032632027838
TransformedTargetRegressor 's R2 :  0.3483720987315203
-----------------------------------------------
0.3483720987315203 0.416032632027838
TweedieRegressor 's R2 :  0.07959167686934143
-----------------------------------------------
0.07959167686934143 0.416032632027838
Best R2 & model & scaler:  0.416032632027838  &  Lasso & MinMaxScaler()

In [11]:

print("Best R2 & model & scaler: ",final_r2," & ",final_r2_model, "&",scaler)

Best R2 & model & scaler:  0.416032632027838  &  Lasso & MinMaxScaler()

이번에는 최종적으로 최선의 r2,model,scaler를 알아보았습니다.¶

실전에서는 위 리스트에 더 다양한 모델과 scaler를 넣어서 최선의 파라미터가 뭔지 쉽게 파악이 가능합니다!¶

In [ ]:

LIST

'scikit-learn(sklearn)' 카테고리의 다른 글

sklearn_iris (0)	2021.07.21
sklearn_wine (0)	2021.07.15
사이킷런 sklearn - boston house price (1)	2021.07.06
Sklearn 함수 (0)	2021.06.28

파이썬 딥러닝

고정 헤더 영역

메뉴 레이어

메뉴 리스트

검색 레이어

검색 영역

상세 컨텐츠

본문 제목

본문

사이킷런(scikit-learn / sklearn) - diabetes datasets¶

필요모듈 import¶

train_test_split 함수를 통해서 해당 데이터셋을 train/test으로 나누겠습니다.¶

이전과 마찬가지로 all_estimator를 이용해서 대략적으로 어느정도의 값이 나오는지 확인해 봅니다.¶

또한 이번에는 standard scaler를 포함시켜 minmax 와 standard scaler중에 누가 더 좋은 성능을 내는지 확인해 봅시다.¶

이번에는 최종적으로 최선의 r2,model,scaler를 알아보았습니다.¶

실전에서는 위 리스트에 더 다양한 모델과 scaler를 넣어서 최선의 파라미터가 뭔지 쉽게 파악이 가능합니다!¶

'scikit-learn(sklearn)' 카테고리의 다른 글

관련글 더보기

댓글 영역

추가 정보

티스토리툴바