Question

我正在尝试用我预处理过的数据集实现k最近邻算法。我将数据导入为pandas数据框，然后将其转换为numpy数组，但发生以下错误-

      File "/home/user/Documents/Mooc_implementation.py", line 8, in <module>
  x = num_data[:,:10]
File "/usr/lib/python2.7/dist-packages/numpy/core/records.py", line 499, in __getitem__
  obj = super(recarray, self).__getitem__(indx)
IndexError: too many indices for array

这是我的代码-

    import numpy as np
    import matplotlib.pyplot as plt
    import pandas as pd

    dataset = pd.read_csv('/home/user/Documents/MOOC dataset cleaned/student_reg_vle_info_assessment.csv')
    num_data = dataset.to_records(index=False)

    x = num_data[:,:10]
    y = num_data[:,10:11]

    from sklearn.model_selection import train_test_split
    X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.20)

    from sklearn.neighbors import KNeighborsClassifier
    classifier = KNeighborsClassifier(n_neighbors=4)
    classifier.fit(X_train, y_train)

    y_pred = classifier.predict(X_test)

    from sklearn.metrics import classification_report, confusion_matrix
    print(confusion_matrix(y_test, y_pred))
    print(classification_report(y_test, y_pred))

我该怎么办？

dataset.head（）-的输出

date_submitted  date_registration  date_unregistration  sum_click  \
0              18               -159                  445         16   
1              22                -53                  445          4   
2              30                -92                   12          3   
3              17                -52                  445          1   
4              26               -176                  445          5   

num_of_prev_attempts  age_band  region  highest_education  studied_credits  \
0                     0         0       0                  0              240   
1                     0         1       1                  0               60   
2                     0         1       2                  1               60   
3                     0         1       3                  1               60   
4                     0         2       4                  2               60   

score  final_result  
0     78             0  
1     70             0  
2     87             2  
3     72             0  
4     69             0  
[Finished in 0.274s]

Answer 1

在您的情况下，为什么要dataset.to_records(index=False)，它将转换为数组，因此不能像num_data[:,:10]那样使用它。并且也无需将dataset.to_records(index=False)转换为train_test_split。

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd

num_data = pd.read_csv('/home/user/Documents/MOOC dataset cleaned/student_reg_vle_info_assessment.csv')
# num_data = dataset.to_records(index=False)

x = num_data.iloc[:,:10]
y = num_data.iloc[:,10:11]

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.20)

from sklearn.neighbors import KNeighborsClassifier
classifier = KNeighborsClassifier(n_neighbors=4)
classifier.fit(X_train, y_train)

y_pred = classifier.predict(X_test)

from sklearn.metrics import classification_report, confusion_matrix
print(confusion_matrix(y_test, y_pred))
print(classification_report(y_test, y_pred))

Answer 2

Q ：为什么会出现此错误？

A ：如果您有熊猫数据集并需要对其进行索引，则需要使用.iloc方法。为问题中的数据集建立索引的方式可以很好地用于numpy数组索引。

使用此：

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd

dataset = pd.read_csv('/home/user/Documents/MOOC dataset cleaned/student_reg_vle_info_assessment.csv')

x = dataset.iloc[:,:10]
y = dataset.iloc[:,10:11]

如何索引（切片）熊猫数据框

2 个答案: