python多标签分类模版

 from sklearn.multioutput import MultiOutputClassifier
 from sklearn.ensemble import RandomForestClassifier
 from sklearn.ensemble import ExtraTreesClassifier
 import numpy as np
 from pandas import read_csv
 import pandas as pd
 
 root1="F:/goverment/shuili2/techproblem_text_train.csv"
 root2="F:/goverment/shuili2/techproblem_text_test.csv"
 root3="F:/goverment/shuili2/text_train_4problem.csv"
 root4="F:/goverment/shuili2/text_test_4problem.csv"
 
 '''大类小类一起预测'''
 #root2="./id="+str(id_num)+"_process_data.csv"
 dataset1 = read_csv(root1) #数据转化为数组
 dataset1=dataset1.values
 dataset2 = read_csv(root2) #数据转化为数组
 dataset2=dataset2.values
 X_train=dataset1[:,:28]# 到28之前都是变量
 Y_train=dataset1[:,28:]# 28到之后都是lable
 X_test=dataset2[:,:28]
 Y_test=dataset2[:,28:]
 
 print('多输出多分类器真实输出分类:\n',Y_train)
 n_samples, n_features = X_train.shape #4000 29
 n_outputs = Y_train.shape[1] # 4000*8
 n_classes = 50 # 每种输出有50种分类
 forest = RandomForestClassifier(n_estimators=500,random_state=1)  # 生成随机森林多分类器
 multi_target_forest = MultiOutputClassifier(forest)  # 构建多输出多分类器
 y_pred = multi_target_forest.fit(X_train, Y_train).predict(X_train)
 print('多输出多分类器预测输出分类:\n',y_pred)
 pp=multi_target_forest.predict(X_test)
 a=pp
 k=0
 for i in range(len(a)):
     if a[i][0]==Y_test[i][0] and a[i][1]==Y_test[i][1] and a[i][2]==Y_test[i][2] and a[i][3]==Y_test[i][3] and a[i][4]==Y_test[i][4] and a[i][5]==Y_test[i][5] and a[i][6]==Y_test[i][6] and a[i][7]==Y_test[i][7]:
         k+=1
 aa=k/1328*1.0
 print(aa)
 
 '''只预测大类'''
 #root2="./id="+str(id_num)+"_process_data.csv"
 dataset3 = read_csv(root1) #数据转化为数组
 dataset3=dataset3.values
 dataset4 = read_csv(root2) #数据转化为数组
 dataset4=dataset4.values
 X_train_big=dataset3[:,:28]
 Y_train_big=dataset3[:,28:32]
 X_test_big=dataset4[:,:28]
 Y_test_big=dataset4[:,28:32]
 print('只预测大类：多输出多分类器真实输出分类:\n',Y_train_big)
 n_samples, n_features = X_train_big.shape #4000 29
 n_outputs = Y_train_big.shape[1] # 4000*8
 n_classes = 11 # 每种输出有11种分类
 forest = RandomForestClassifier(n_estimators=200,random_state=1)  # 生成随机森林多分类器
 multi_target_forest = MultiOutputClassifier(forest)  # 构建多输出多分类器
 y_pred = multi_target_forest.fit(X_train_big, Y_train_big).predict(X_train_big)
 print('多输出多分类器预测输出分类:\n',y_pred)
 pp=multi_target_forest.predict(X_test_big)
 a=pp
 k=0
 for i in range(len(a)):
     if a[i][0]==Y_test_big[i][0] and a[i][1]==Y_test_big[i][1] and a[i][2]==Y_test_big[i][2] and a[i][3]==Y_test_big[i][3]:
         k+=1
 aa=k/1328*1.0
 print(aa)
 
 '''只预测小类'''
 #root2="./id="+str(id_num)+"_process_data.csv"
 dataset4 = read_csv(root3) #数据转化为数组
 dataset4=dataset4.values
 dataset5 = read_csv(root4) #数据转化为数组
 dataset5=dataset5.values
 X_train_samll=dataset4[:,:28]
 Y_train_samll=dataset4[:,28:32]
 X_test_samll=dataset5[:,:28]
 Y_test_samll=dataset5[:,28:32]
 print('只预测小类：多输出多分类器真实输出分类:\n',Y_train_samll)
 n_samples, n_features = X_train_samll.shape #4000 29
 n_outputs = Y_train_samll.shape[1] # 4000*4
 n_classes = 61 # 每种输出有61种分类
 forest = RandomForestClassifier(n_estimators=200,random_state=1)  # 生成随机森林多分类器
 multi_target_forest = MultiOutputClassifier(forest)  # 构建多输出多分类器
 y_pred = multi_target_forest.fit(X_train_samll, Y_train_samll).predict(X_train_samll)
 print('多输出多分类器预测输出分类:\n',y_pred)
 pp=multi_target_forest.predict(X_test_samll)
 a=pp
 k=0
 for i in range(len(a)):
     if a[i][0]==Y_test_samll[i][0] and a[i][1]==Y_test_samll[i][1] and a[i][2]==Y_test_samll[i][2] and a[i][3]==Y_test_samll[i][3]:
         k+=1
 aa=k/1328*1.0
 print(aa)
 
 '''
 from pandas import read_csv
 import pandas as pd
 import numpy as np
 from skmultilearn.problem_transform import BinaryRelevance
 from sklearn.naive_bayes import GaussianNB
 from sklearn.metrics import accuracy_score
 
 root1="D:/Anaconda3-5.0.1-Windows-x86_64/anaconda/work/shuili2/data.csv"
 #root2="./id="+str(id_num)+"_process_data.csv"
 dataset = read_csv(root1) #数据转化为数组
 dataset=dataset.values
 x_train=dataset[:4000,:29]
 y_train=dataset[:4000,29:]
 
 x_test=dataset[4000:,:29]
 y_test=dataset[4000:,29:]
 
 # initialize binary relevance multi-label classifier
 # with a gaussian naive bayes base classifier
 classifier = BinaryRelevance(GaussianNB())
 
 # train
 classifier.fit(x_train, y_train)
 
 # predict
 predictions = classifier.predict(x_test)
 accuracy_score(y_test,predictions)
 '''
 
 '''---------------------------------'''
 '''
 import numpy as np
 import pandas as pd
 from keras.models import Sequential
 from keras.layers import Dense, Dropout
 from keras.wrappers.scikit_learn import KerasClassifier
 from keras.utils import np_utils
 from sklearn.model_selection import train_test_split, KFold, cross_val_score
 from sklearn.preprocessing import LabelEncoder
 from pandas import read_csv
 from sklearn.naive_bayes import GaussianNB
 from sklearn.metrics import accuracy_score
 
 root1="D:/Anaconda3-5.0.1-Windows-x86_64/anaconda/work/shuili2/data.csv"
 #root2="./id="+str(id_num)+"_process_data.csv"
 dataset = read_csv(root1) #数据转化为数组
 dataset=dataset.values
 
 # load dataset
 dataframe = pd.read_csv("data.csv", header=None)
 dataset = dataframe.values
 X = dataset[:, 0:29].astype(float)
 Y = dataset[:, 29:]
 
 # encode class values as integers
 #encoder = LabelEncoder()
 #encoded_Y = encoder.fit_transform(Y)
 # convert integers to dummy variables (one hot encoding)
 #dummy_y = np_utils.to_categorical(encoded_Y)
 
 # define model structure
 def baseline_model():
     model = Sequential()
     model.add(Dense(output_dim=10, input_dim=29, activation='relu'))
     model.add(Dropout(0.2))
     model.add(Dense(output_dim=8, input_dim=10, activation='softmax'))
     # Compile model
     model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])
     return model
 estimator = KerasClassifier(build_fn=baseline_model, nb_epoch=200, batch_size=50)
 # splitting data into training set and test set. If random_state is set to an integer, the split datasets are fixed.
 X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.01, random_state=0)
 estimator.fit(X_train, Y_train)
 
 # make predictions
 pred = estimator.predict(X_test)
 
 # inverse numeric variables to initial categorical labels
 #init_lables = encoder.inverse_transform(pred)
 
 # k-fold cross-validate
 seed = 42
 np.random.seed(seed)
 kfold = KFold(n_splits=10, shuffle=True, random_state=seed)
 results = cross_val_score(estimator, X, Y, cv=kfold)
 '''

from pandas import read_csv
 
root1="F:/goverment/shuili2/techproblem_text_train.csv"
root2="F:/goverment/shuili2/techproblem_text_test.csv"
root3="F:/goverment/shuili2/text_train_4problem.csv"
root4="F:/goverment/shuili2/text_test_4problem.csv"
 
'''大类小类一起预测'''
#root2="./id="+str(id_num)+"_process_data.csv"
dataset1 = read_csv(root1) #数据转化为数组
dataset1=dataset1.values
dataset2 = read_csv(root2) #数据转化为数组
dataset2=dataset2.values
X_train=dataset1[:,:28]
Y_train=dataset1[:,28:]
X_test=dataset2[:,:28]
Y_test=dataset2[:,28:]
 
from pprint import pprint
pprint(dataset1)
 
#＃使用二进制相关性
#scikit-multilearn
from skmultilearn.problem_transform import BinaryRelevance
from sklearn.naive_bayes import GaussianNB
 
#initialize二进制相关多标签分类器
#用高斯朴素贝叶斯基分类器
classifier = BinaryRelevance(GaussianNB())
 
 #训练
classifier.fit(X_train, Y_train)
 
 #预测
predictions = classifier.predict(X_test)
 
#计算精度用
from sklearn.metrics import accuracy_score
accuracy_score(Y_test,predictions)