Question

我很困惑为什么我的输入形状输入错误。失败的行是

test_datapoint_encoded[i] = int(label_encoder[count].transform(test_datapoint[i]))

我在这里读到，显然转换函数不适用于列表，但我使用了一个不同的例子，它可以正常工作。 test_datapoint和test_datapoint_encoded的形状相同。我尝试使用numpy数组，但我得到了同样的错误。

import numpy as np
import matplotlib.pyplot as plt
from sklearn.metrics import classification_report, mean_absolute_error
from sklearn import cross_validation, preprocessing
from sklearn.ensemble import ExtraTreesRegressor
from sklearn.metrics import classification_report

# Load input data
input_file = 'traffic_data.txt'
data = []

with open(input_file, 'r') as f:
    for line in f.readlines():
        items = line[:-1].split(',')
        data.append(items)

data = np.array(data)    

    #convert string to numerical
    label_encoder =[]
    X_encoded = np.empty(data.shape)

    for i, item in enumerate(data[0]):
        if item.isdigit():
            X_encoded[:,i] = data[:,i]
        else:
            label_encoder.append(preprocessing.LabelEncoder())
            X_encoded[:,i] = label_encoder[-1].fit_transform(data[:,i])


    #evrything but the last column        
    X = X_encoded[:, :-1].astype(int)

    #only the last collumn
    y = X_encoded[:, -1].astype(int)        

    #split into test and train
    X_train, X_test, y_train, y_test = cross_validation.train_test_split(X, y, test_size = 0.25, random_state=5)

    #create extreme forest regressor
    params = {'n_estimators': 100, 'max_depth': 4, 'random_state': 0}
    ext_regressor = ExtraTreesRegressor(**params)
    ext_regressor.fit(X_train, y_train)


          # Compute the regressor performance on test data
    y_pred = ext_regressor.predict(X_test)
    print("Mean absolute error:", round(mean_absolute_error(y_test, y_pred),
    2))  

        enter code here

    # Testing encoding on single data instance
    test_datapoint = ['Saturday', '10:20', 'Atlanta', 'no']
    test_datapoint_encoded = [0, 0, 0, 0]


    count = 0
    for i, item in enumerate(test_datapoint):
        if item.isdigit():
            test_datapoint_encoded[i] = int(test_datapoint[i])
        else:
            test_datapoint_encoded[i] = int(label_encoder[count].transform(test_datapoint[i]))
            count = count + 1

    test_datapoint_encoded = np.array(test_datapoint_encoded)


    # Predict the output for the test datapoint
    print("Predicted traffic:",
    int(ext_regressor.predict([test_datapoint_encoded])[0]))

我使用了一个更简单的标签编码器示例，之前没有任何错误。

import numpy as np
from sklearn import preprocessing

#definde sample labels
input_labels = ['red', 'black', 'red', 'green', 'black', 'yellow', 'white']


#creating label encoder and train it.
encoder = preprocessing.LabelEncoder()
#train my encoder to associate each color with a digit
encoder.fit(input_labels)

#print the mapping between words and numbers
print("\nLabel mapping:")
for i, item in enumerate(encoder.classes_):
    print(item, '--->', i)

#encode a set of test labels and compare results
test_labels = ['green', 'red', 'black']

encoded_values = encoder.transform(test_labels)
print("\nLabels =", test_labels)
print("Encoded values =", list(encoded_values))


##ii can now decode the numbers into colors
encoded_test_values = [3, 0 ,4, 1]
decoded_list = encoder.inverse_transform(encoded_test_values)
print("\nEncoded values =", encoded_test_values)
print("Decoded labels =", list(decoded_list))

Answer 1

在你的行

 X_encoded[:,i] = label_encoder[-1].fit_transform(data[:,i])

您正在使用N维数组输入训练标签编码器。但在你的行

 test_datapoint_encoded[i] = int(label_encoder[count].transform(test_datapoint[i]))

传递给标签编码器的值并不完全相同，因此您可能会收到“输入形状错误”

输入形状不好的LabelEncoder

1 个答案: