python - 至少指定一个标签必须在 y_true 中

Question

我想根据y_test和pred_test得到一个混淆矩阵，但是提出了一个问题“At least one label specified must be in y_true”，我不知道为什么

metrics.confusion_matrix(np.argmax(y_test,axis=1),pred_test)


  y_test =  [[0. 1. 0. 0. 0. 0.]
     [0. 0. 0. 0. 0. 1.]
     [0. 0. 0. 0. 1. 0.]
     ...
     [0. 0. 0. 1. 0. 0.]
     [0. 0. 1. 0. 0. 0.]
     [0. 0. 1. 0. 0. 0.]]

   pred_test = [1 4 5 ... 3 2 2]
   np.argmax(y_test,axis=1) = [1 5 4 ... 3 2 2]

  File "D:\Anaconda\lib\site-packages\sklearn\metrics\classification.py", line 259, in confusion_matrix
    raise ValueError("At least one label specified must be in y_true")
ValueError: At least one label specified must be in y_true

我创建了一个卷积神经网络。模型并使用交叉验证进行估计，最终生成混淆矩阵。现在生成混淆矩阵有问题。

数据集是enter link description here .完整代码如下:

 import matplotlib
    #matplotlib.use('Agg')
    import timing
    from keras.layers import Input,Dense,Conv2D,MaxPooling2D,UpSampling2D,Flatten
    from keras.models import Model
    from keras import backend as K
    from keras.utils.np_utils import to_categorical
    import numpy as np
    import pandas as pd
    import seaborn as sns
    from keras.models import Sequential# 导入Sequential
    from keras.utils import np_utils, generic_utils
    from keras.callbacks import LearningRateScheduler
    import os
    from keras.layers import Dropout
    from keras.backend.tensorflow_backend import set_session
    import tensorflow as tf
    from sklearn.model_selection import train_test_split,  cross_val_score
    from sklearn.cross_validation import KFold, StratifiedKFold
    from keras.wrappers.scikit_learn import KerasClassifier
    from sklearn.preprocessing import LabelEncoder
    from sklearn import metrics
    import time
    from scipy import stats
    from keras import optimizers
    import matplotlib.pyplot as plt
    from keras import regularizers
    import keras
    from keras.callbacks import TensorBoard
    config = tf.ConfigProto(allow_soft_placement=True)
    gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=0.9)
    config.gpu_options.allow_growth = True
    sess = tf.Session(config=config)
    os.environ['TF_CPP_MIN_LOG_LEVEL']='2'
    time1 = time.time()
    class LossHistory(keras.callbacks.Callback):
        def on_train_begin(self, logs={}):
            self.losses = {'batch':[], 'epoch':[]}
            self.accuracy = {'batch':[], 'epoch':[]}
            self.val_loss = {'batch':[], 'epoch':[]}
            self.val_acc = {'batch':[], 'epoch':[]}

        def on_batch_end(self, batch, logs={}):
            self.losses['batch'].append(logs.get('loss'))
            self.accuracy['batch'].append(logs.get('acc'))
            self.val_loss['batch'].append(logs.get('val_loss'))
            self.val_acc['batch'].append(logs.get('val_acc'))

        def on_epoch_end(self, batch, logs={}):
            self.losses['epoch'].append(logs.get('loss'))
            self.accuracy['epoch'].append(logs.get('acc'))
            self.val_loss['epoch'].append(logs.get('val_loss'))
            self.val_acc['epoch'].append(logs.get('val_acc'))

        def loss_plot(self, loss_type):
            iters = range(len(self.losses[loss_type]))
            plt.figure()
            # acc
            plt.plot(iters, self.accuracy[loss_type], 'r', label='train acc')
            # loss
            plt.plot(iters, self.losses[loss_type], 'g', label='train loss')
            if loss_type == 'epoch':
                # val_acc
                plt.plot(iters, self.val_acc[loss_type], 'b', label='val acc')
                # val_loss
                plt.plot(iters, self.val_loss[loss_type], 'k', label='val loss')
            plt.grid(True)
            plt.xlabel(loss_type)
            plt.ylabel('acc-loss')
            plt.legend(loc="center")
            plt.show()
            #plt.savefig('common.png')


    #dataset
    RANDOM_SEED = 42
    def read_data(file_path):
        column_names = ['user-id', 'activity', 'timestamp', 'x-axis', 'y-axis', 'z-axis']
        m = pd.read_csv(file_path,names=column_names, header=None,sep=',')
        return m
    def feature_normalize(dataset):
        mu = np.mean(dataset,axis=0)
        sigma = np.std(dataset,axis=0)
        return (dataset-mu)/sigma

    dataset1 = read_data('ab.txt')
    dataset = pd.DataFrame(dataset1)
    dataset['x-axis'] = feature_normalize(dataset['x-axis'])
    dataset['y-axis'] = feature_normalize(dataset['y-axis'])
    dataset['z-axis'] = feature_normalize(dataset['z-axis'])

    N_TIME_STEPS = 200
    N_FEATURES = 3
    step = 200
    segments = []
    labels = []
    for i in range(0, len(dataset) - N_TIME_STEPS, step):
        xs = dataset['x-axis'].values[i: i + N_TIME_STEPS]
        ys = dataset['y-axis'].values[i: i + N_TIME_STEPS]
        zs = dataset['z-axis'].values[i: i + N_TIME_STEPS]
        label = stats.mode(dataset['activity'][i: i + N_TIME_STEPS])[0][0]
        segments.append([xs, ys, zs])
        labels.append(label)
    print("reduced size of data", np.array(segments).shape)
    reshaped_segments = np.asarray(segments,dtype=np.float32).reshape(-1,1, N_TIME_STEPS, 3)
    print("Reshape the segments", np.array(reshaped_segments).shape)
    #x_train1, x_val_test, y_train1, y_val_test = train_test_split(reshaped_segments, labels, test_size=0.25, random_state=RANDOM_SEED)

    batch_size = 128     
    num_classes =6

    def create_model():
        input_shape = Input(shape=(1,200,3))
        x = Conv2D(5, kernel_size=(1, 1), padding='valid')(input_shape)
        x1 = keras.layers.concatenate([input_shape, x], axis=-1)

        x = Conv2D(50, kernel_size=(1, 7),padding='valid',
                     kernel_initializer='glorot_uniform',
        kernel_regularizer = keras.regularizers.l2(0.0015))(x1)


        x = keras.layers.core.Activation('relu')(x)
        x = MaxPooling2D(pool_size=(1, 2))(x)
        x = Conv2D(50, kernel_size=(1, 7),padding='valid',kernel_initializer='glorot_uniform',
               kernel_regularizer=keras.regularizers.l2(0.0015))(x)
        x = keras.layers.core.Activation('relu')(x)
        x = MaxPooling2D(pool_size=(1, 2))(x)

        x = Flatten()(x)
        x = Dropout(0.9)(x)
        output = Dense(num_classes, activation='softmax',kernel_initializer='glorot_uniform',)(x)
        model = Model(inputs=input_shape,outputs=output)
        model.summary()

        sgd = optimizers.SGD(lr=0.005,decay=1e-6,momentum=0.9,nesterov=True)
        model.compile(loss=keras.losses.categorical_crossentropy,
                  optimizer=sgd,
                  metrics=['accuracy'])
        return model
    history = LossHistory()
    epochs = 4000


    #setting learning rate
    def scheduler(epoch):
        if epoch > 0.75 * epochs:
            lr = 0.0005
        elif epoch > 0.25 * epochs:
            lr = 0.001
        else:
            lr = 0.005
        return lr

    scheduler = LearningRateScheduler(scheduler)
    estimator = KerasClassifier(build_fn=create_model)
    #divide dataset

    scores = []
    confusions = []   
    sign = ['DOWNSTAIRS','JOGGING','SITTING','STANDING','UPSTAIRS','WALKING']
    encoder = LabelEncoder()
    encoder_y = encoder.fit_transform(labels)
    train_labels = to_categorical(encoder_y,num_classes=None)

    #kfold = StratifiedKFold(reshaped_segments.shape[0],n_folds=10,shuffle=True,random_state=42)
    kfold = StratifiedKFold(labels,n_folds=3,shuffle=True,random_state=42)
    for train_index,test_index in kfold:
        print(test_index)
        x_train, x_test = reshaped_segments[train_index], reshaped_segments[test_index]
        y_train, y_test = train_labels[train_index], train_labels[test_index]
        estimator.fit(x_train,y_train,callbacks=[scheduler,history],epochs=10,batch_size=128,verbose=0)
        scores.append(estimator.score(x_test,y_test))
        print(y_test)
        print(type(y_test))
        pred_test = estimator.predict(x_test)  
        print(pred_test)
        print(np.argmax(y_test,axis=1))
        confusions.append(metrics.confusion_matrix(np.argmax(y_test,axis=1),pred_test,sign))

    matrix = [[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]]

    for i in np.arange(n_folds-1):
        for j in len(confusions[0]):
            for k in len(confusions[0][0]):
                matrix[j][k] = matrix[j][k] + confusions[i][j][k] + confusions[i+1][j][k]  

    model.save('model.h5')  
    model.save_weights('my_model_weights.h5')
    print('score:',scores)
    scores = np.mean(scores)
    print('mean:',scores)

    plt.figure(figsize=(16,14))     
    sns.heatmap(matrix, xticklabels=sign, yticklabels=sign, annot=True, fmt="d");
    plt.title("CONFUSION MATRIX : ")
    plt.ylabel('True Label')
    plt.xlabel('Predicted label')
    plt.savefig('cmatrix.png')
    plt.show();

Answer 1

错误不在您的主代码中，而是在符号的定义中。当您将符号定义为

 sign = ['DOWNSTAIRS','JOGGING','SITTING','STANDING','UPSTAIRS','WALKING']

系统无法读取您的标签，因为它正在寻找标签 0、1、2、3、4、5 作为错误试图说明的内容，即它在 y_pred 中找不到任何标签。更改符号为

 sign = [1,2,3,4,5]

应该修复错误。至于你现在做什么，它相当简单，只需将你的结果映射为这个数组，然后在实际预测(部署)期间，只需换出标签的数值。