sklearn中报错ValueError: Expected 2D array, got 1D array instead:

Posted 2021-02-14 --wwwww

 

 1 from sklearn.linear_model import LinearRegression
 2 lr = LinearRegression()
 3 print(tr_x.shape,tr_y.shape)
 4 lr.fit(tr_x,tr_y)
 5 
 6 
 7 # 报错
 8 (64,) (64,)
 9 Traceback (most recent call last):
10   File "F:/Python_Project/sklearn2_2/zong_fu_xi/A_02.py", line 51, in <module>
11     lr.fit(tr_x,tr_y)
12   File "F:Python_Projectmachine_learning_project_01libsite-packagessklearnlinear_modelase.py", line 458, in fit
13     y_numeric=True, multi_output=True)
14   File "F:Python_Projectmachine_learning_project_01libsite-packagessklearnutilsvalidation.py", line 756, in check_X_y
15     estimator=estimator)
16   File "F:Python_Projectmachine_learning_project_01libsite-packagessklearnutilsvalidation.py", line 552, in check_array
17     "if it contains a single sample.".format(array))
18 ValueError: Expected 2D array, got 1D array instead:
19 array=[ 9.1802  5.8707  7.4239 13.176   7.0708  5.6397 18.959   5.0269  8.5186
20  21.279   5.7737 11.708   8.3829  6.3654  6.4296  6.8825  6.3534  7.4764
21   5.5204  8.8254  5.5277  7.9334 22.203   5.3077  5.734   8.0959  5.5649
22   7.6031 14.164   9.2482  5.7077  9.3102  5.0365  5.8918  9.7687  5.3794
23   6.5479  6.1891  5.2524  7.5402  8.2934 13.394  10.136  20.27    7.6366
24   7.2259 10.274  12.828  12.836   5.8014  5.4069  8.2951  9.4536  8.4084
25   7.3345  5.6063  5.4901  6.5159  5.7107  5.3054  5.4994  7.2182 11.7
26   7.0931].
27 Reshape your data either using array.reshape(-1, 1) if your data has a single feature or array.reshape(1, -1) if it contains a single sample.

# 造成错误原因：

这是由于在新版的sklearn中，所有的数据都应该是二维矩阵，哪怕它只是单独一行或一列（比如前面做预测时，仅仅只用了一个样本数据），前面程序第3行输出的维度是（64，）一维的，所以需要使用.reshape(1,-1)进行转换，具体操作如下。

需改为

from sklearn.linear_model import LinearRegression
lr = LinearRegression()
tr_x = np.array(tr_x).reshape(1,-1)
te_x = np.array(te_x).reshape(1,-1)
tr_y = np.array(tr_y).reshape(1,-1)
te_y = np.array(te_y).reshape(1,-1)
print(tr_x.shape,tr_y.shape)
lr.fit(tr_x,tr_y)

此时这个错误就解决了