from sklearn import datasets#引入數(shù)據(jù)集,sklearn包含眾多數(shù)據(jù)集
from sklearn.model_selection import train_test_split#將數(shù)據(jù)分為測(cè)試集和訓(xùn)練集
from sklearn.neighbors import KNeighborsClassifier#利用鄰近點(diǎn)方式訓(xùn)練數(shù)據(jù)
###引入數(shù)據(jù)###
iris=datasets.load_iris()#引入iris鳶尾花數(shù)據(jù),iris數(shù)據(jù)包含4個(gè)特征變量
iris_X=iris.data#特征變量
iris_y=iris.target#目標(biāo)值
X_train,X_test,y_train,y_test=train_test_split(iris_X,iris_y,test_size=0.3)#利用train_test_split進(jìn)行將訓(xùn)練集和測(cè)試集進(jìn)行分開(kāi)，test_size占30%
print(y_train)#我們看到訓(xùn)練數(shù)據(jù)的特征值分為3類
'''
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2
 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
 2 2]
 '''
###訓(xùn)練數(shù)據(jù)###
knn=KNeighborsClassifier()#引入訓(xùn)練方法
knn.fit(X_train,y_train)#進(jìn)行填充測(cè)試數(shù)據(jù)進(jìn)行訓(xùn)練
###預(yù)測(cè)數(shù)據(jù)###
print(knn.predict(X_test))#預(yù)測(cè)特征值
'''
[1 1 1 0 2 2 1 1 1 0 0 0 2 2 0 1 2 2 0 1 0 0 0 0 0 0 2 1 0 0 0 1 0 2 0 2 0
 1 2 1 0 0 1 0 2]
'''
print(y_test)#真實(shí)特征值
'''
[1 1 1 0 1 2 1 1 1 0 0 0 2 2 0 1 2 2 0 1 0 0 0 0 0 0 2 1 0 0 0 1 0 2 0 2 0
 1 2 1 0 0 1 0 2]
'''

from sklearn import datasets#引入數(shù)據(jù)集
#構(gòu)造的各種參數(shù)可以根據(jù)自己需要調(diào)整
X,y=datasets.make_regression(n_samples=100,n_features=1,n_targets=1,noise=1)
###繪制構(gòu)造的數(shù)據(jù)###
import matplotlib.pyplot as plt
plt.figure()
plt.scatter(X,y)
plt.show()

from sklearn import datasets
from sklearn.linear_model import LinearRegression#引入線性回歸模型
###引入數(shù)據(jù)###
load_data=datasets.load_boston()
data_X=load_data.data
data_y=load_data.target
print(data_X.shape)
#(506, 13)data_X共13個(gè)特征變量
###訓(xùn)練數(shù)據(jù)###
model=LinearRegression()
model.fit(data_X,data_y)
model.predict(data_X[:4,:])#預(yù)測(cè)前4個(gè)數(shù)據(jù)
###屬性和功能###
print(model.coef_)
'''
[ -1.07170557e-01   4.63952195e-02   2.08602395e-02   2.68856140e+00
  -1.77957587e+01   3.80475246e+00   7.51061703e-04  -1.47575880e+00
   3.05655038e-01  -1.23293463e-02  -9.53463555e-01   9.39251272e-03
  -5.25466633e-01]
'''
print(model.intercept_)
#36.4911032804
print(model.get_params())#得到模型的參數(shù)
#{'copy_X': True, 'normalize': False, 'n_jobs': 1, 'fit_intercept': True}
print(model.score(data_X,data_y))#對(duì)訓(xùn)練情況進(jìn)行打分
#0.740607742865
 

from sklearn import preprocessing
import numpy as np
a=np.array([[10,2.7,3.6],
            [-100,5,-2],
            [120,20,40]],dtype=np.float64)
print(a)
print(preprocessing.scale(a))#將值的相差度減小
'''
[[  10\.     2.7    3.6]
 [-100\.     5\.    -2\. ]
 [ 120\.    20\.    40
[[ 0\.         -0.85170713 -0.55138018]
 [-1.22474487 -0.55187146 -0.852133  ]
 [ 1.22474487  1.40357859  1.40351318]]
'''
 

from sklearn.model_selection import train_test_split
from sklearn.datasets.samples_generator import make_classification
from sklearn.svm import SVC
import matplotlib.pyplot as plt
###生成的數(shù)據(jù)如下圖所示###
plt.figure
X,y=make_classification(n_samples=300,n_features=2,n_redundant=0,n_informative=2,             random_state=22,n_clusters_per_class=1,scale=100)
plt.scatter(X[:,0],X[:,1],c=y)
plt.show()
###利用minmax方式對(duì)數(shù)據(jù)進(jìn)行規(guī)范化###
X=preprocessing.minmax_scale(X)#feature_range=(-1,1)可設(shè)置重置范圍
X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.3)
clf=SVC()
clf.fit(X_train,y_train)
print(clf.score(X_test,y_test))
#0.933333333333
#沒(méi)有規(guī)范化之前我們的訓(xùn)練分?jǐn)?shù)為0.511111111111,規(guī)范化后為0.933333333333,準(zhǔn)確度有很大提升

from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
###引入數(shù)據(jù)###
iris=load_iris()
X=iris.data
y=iris.target
###訓(xùn)練數(shù)據(jù)###
X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.3)
#引入交叉驗(yàn)證,數(shù)據(jù)分為5組進(jìn)行訓(xùn)練
from sklearn.model_selection import cross_val_score
knn=KNeighborsClassifier(n_neighbors=5)#選擇鄰近的5個(gè)點(diǎn)
scores=cross_val_score(knn,X,y,cv=5,scoring='accuracy')#評(píng)分方式為accuracy
print(scores)#每組的評(píng)分結(jié)果
#[ 0.96666667  1\.          0.93333333  0.96666667  1\.        ]5組數(shù)據(jù)
print(scores.mean())#平均評(píng)分結(jié)果
#0.973333333333

from sklearn import datasets
from sklearn.model_selection import train_test_split
from sklearn.neighbors import KNeighborsClassifier
from sklearn.model_selection import cross_val_score#引入交叉驗(yàn)證
import  matplotlib.pyplot as plt
###引入數(shù)據(jù)###
iris=datasets.load_iris()
X=iris.data
y=iris.target
###設(shè)置n_neighbors的值為1到30,通過(guò)繪圖來(lái)看訓(xùn)練分?jǐn)?shù)###
k_range=range(1,31)
k_score=[]
for k in k_range:
    knn=KNeighborsClassifier(n_neighbors=k)
    scores=cross_val_score(knn,X,y,cv=10,scoring='accuracy')#for classfication
    k_score.append(loss.mean())
plt.figure()
plt.plot(k_range,k_score)
plt.xlabel('Value of k for KNN')
plt.ylabel('CrossValidation accuracy')
plt.show()
#K過(guò)大會(huì)帶來(lái)過(guò)擬合問(wèn)題,我們可以選擇12-18之間的值

for k in k_range:
    knn=KNeighborsClassifier(n_neighbors=k)
    loss=-cross_val_score(knn,X,y,cv=10,scoring='neg_mean_squared_error')# for regression
    k_score.append(loss.mean())

from sklearn.model_selection import learning_curve
from sklearn.datasets import load_digits
from sklearn.svm import SVC
import matplotlib.pyplot as plt
import numpy as np
#引入數(shù)據(jù)
digits=load_digits()
X=digits.data
y=digits.target
#train_size表示記錄學(xué)習(xí)過(guò)程中的某一步,比如在10%,25%...的過(guò)程中記錄一下
train_size,train_loss,test_loss=learning_curve(
    SVC(gamma=0.1),X,y,cv=10,scoring='neg_mean_squared_error',
    train_sizes=[0.1,0.25,0.5,0.75,1]
)
train_loss_mean=-np.mean(train_loss,axis=1)
test_loss_mean=-np.mean(test_loss,axis=1)
plt.figure()
#將每一步進(jìn)行打印出來(lái)
plt.plot(train_size,train_loss_mean,'o-',color='r',label='Training')
plt.plot(train_size,test_loss_mean,'o-',color='g',label='Cross-validation')
plt.legend('best')
plt.show()

from sklearn.model_selection import  validation_curve#將learning_curve改為validation_curve
from sklearn.datasets import load_digits
from sklearn.svm import SVC
import matplotlib.pyplot as plt
import numpy as np
#引入數(shù)據(jù)
digits=load_digits()
X=digits.data
y=digits.target
#改變param來(lái)觀察Loss函數(shù)情況
param_range=np.logspace(-6,-2.3,5)
train_loss,test_loss=validation_curve(
    SVC(),X,y,param_name='gamma',param_range=param_range,cv=10,
    scoring='neg_mean_squared_error'
)
train_loss_mean=-np.mean(train_loss,axis=1)
test_loss_mean=-np.mean(test_loss,axis=1)
plt.figure()
plt.plot(param_range,train_loss_mean,'o-',color='r',label='Training')
plt.plot(param_range,test_loss_mean,'o-',color='g',label='Cross-validation')
plt.xlabel('gamma')
plt.ylabel('loss')
plt.legend(loc='best')
plt.show()

from sklearn import svm
from sklearn import datasets
#引入和訓(xùn)練數(shù)據(jù)
iris=datasets.load_iris()
X,y=iris.data,iris.target
clf=svm.SVC()
clf.fit(X,y)
#引入sklearn中自帶的保存模塊
from sklearn.externals import joblib
#保存model
joblib.dump(clf,'sklearn_save/clf.pkl')
#重新加載model，只有保存一次后才能加載model
clf3=joblib.load('sklearn_save/clf.pkl')
print(clf3.predict(X[0:1]))
#存放model能夠更快的獲得以前的結(jié)果