Skip to main content
 首页 » 编程设计

python之Scikit-learn分类测试,ValueError : setting an array element with a sequence

2025年05月04日9zlslch

使用 tutorial on multiclass adaboost ,我正在尝试对一些具有两个类别的图像进行分类(但我不认为如果问题是二元的,则算法不应该工作)。然后我将扩展我的示例以包括其他类。

我现在的测试很小,一共只有17张图片,10张训练,7张测试。

现在我有两个类:0:没有车辆,1:有车辆 我使用整数标签是因为根据上面链接中的示例,训练数据由基于整数的标签组成。

我只对提供的示例进行了一些编辑,以包含我自己的图像文件,但出现错误。

Traceback (most recent call last): 
  File "C:\Users\app\Documents\Python Scripts\carclassify.py", line 66, in <module> 
    bdt_discrete.fit(X_train, y_train) 
  File "C:\Users\app\Anaconda\lib\site-packages\sklearn\ensemble\weight_boosting.py", line 389, in fit 
    return super(AdaBoostClassifier, self).fit(X, y, sample_weight) 
  File "C:\Users\app\Anaconda\lib\site-packages\sklearn\ensemble\weight_boosting.py", line 99, in fit 
    X = np.ascontiguousarray(array2d(X), dtype=DTYPE) 
  File "C:\Users\app\Anaconda\lib\site-packages\numpy\core\numeric.py", line 408, in ascontiguousarray 
    return array(a, dtype, copy=False, order='C', ndmin=1) 
ValueError: setting an array element with a sequence. 

以下是我的代码,改编自 scikit-learn 网站上的示例:

f = open("PATH_TO_SAMPLES\\samples.txt",'r') 
out = f.read().splitlines() 
import numpy as np 
 
imgs = [] 
tmp_hogs = [] 
# 13 of the images are with vehicles, 4 are without 
labels = [1,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0] 
 
for file in out: 
        filepath = "C:\PATH_TO_SAMPLE_IMAGES\\" + file 
        curr_img = color.rgb2gray(io.imread(filepath)) 
        imgs.append(resize(curr_img,(60,40))) 
        fd, hog_image = hog(curr_img, orientations=8, pixels_per_cell=(16, 16), 
                 cells_per_block=(1, 1), visualise=True) 
        tmp_hogs.append(fd)  
 
img_hogs = np.array(tmp_hogs) 
n_split = 10 
X_train, X_test = img_hogs[:n_split], X[n_split:] # all first ten images with vehicles 
y_train, y_test = labels[:n_split], labels[n_split:] # 3 images with vehicles, 4 without 
 
#now all the code below is straight off the example on scikit-learn's website 
 
bdt_real = AdaBoostClassifier( 
    DecisionTreeClassifier(max_depth=2), 
    n_estimators=600, 
    learning_rate=1) 
 
bdt_discrete = AdaBoostClassifier( 
    DecisionTreeClassifier(max_depth=2), 
    n_estimators=600, 
    learning_rate=1.5, 
    algorithm="SAMME") 
 
bdt_real.fit(X_train, y_train) 
bdt_discrete.fit(X_train, y_train) 
 
real_test_errors = [] 
discrete_test_errors = [] 
 
for real_test_predict, discrete_train_predict in zip( 
        bdt_real.staged_predict(X_test), bdt_discrete.staged_predict(X_test)): 
    real_test_errors.append( 
        1. - accuracy_score(real_test_predict, y_test)) 
    discrete_test_errors.append( 
        1. - accuracy_score(discrete_train_predict, y_test)) 
 
n_trees = xrange(1, len(bdt_discrete) + 1) 
 
pl.figure(figsize=(15, 5)) 
 
pl.subplot(131) 
pl.plot(n_trees, discrete_test_errors, c='black', label='SAMME') 
pl.plot(n_trees, real_test_errors, c='black', 
        linestyle='dashed', label='SAMME.R') 
pl.legend() 
pl.ylim(0.18, 0.62) 
pl.ylabel('Test Error') 
pl.xlabel('Number of Trees') 
 
pl.subplot(132) 
pl.plot(n_trees, bdt_discrete.estimator_errors_, "b", label='SAMME', alpha=.5) 
pl.plot(n_trees, bdt_real.estimator_errors_, "r", label='SAMME.R', alpha=.5) 
pl.legend() 
pl.ylabel('Error') 
pl.xlabel('Number of Trees') 
pl.ylim((.2, 
        max(bdt_real.estimator_errors_.max(), 
            bdt_discrete.estimator_errors_.max()) * 1.2)) 
pl.xlim((-20, len(bdt_discrete) + 20)) 
 
pl.subplot(133) 
pl.plot(n_trees, bdt_discrete.estimator_weights_, "b", label='SAMME') 
pl.legend() 
pl.ylabel('Weight') 
pl.xlabel('Number of Trees') 
pl.ylim((0, bdt_discrete.estimator_weights_.max() * 1.2)) 
pl.xlim((-20, len(bdt_discrete) + 20)) 
 
# prevent overlapping y-axis labels 
pl.subplots_adjust(wspace=0.25) 
pl.show() 

编辑

我输入了

print tmp_hogs 

输出是这样的:

[ array([ 0.27621208,  0.11038658,  0.10698133, ...,  0.08661556,        0.04612063,  0.0280782 ]),  
        array([  0.00000000e+00,   0.00000000e+00,   0.00000000e+00, ..., -1.29909838e-15,  -7.01780982e-17,  -1.24900943e-15]),  
        array([ 0.0503603 ,  0.1497235 ,  0.2372957 , ...,  0.07249325, 0.04545541,  0.00903818]),  
        array([ 0.27299191,  0.13122109,  0.0719268 , ...,  0.0848522 ,  0.04789403,  0.01387038]),  
        array([  0.00000000e+00,   0.00000000e+00,   0.00000000e+00, ...,  3.32140617e-17,  -6.58924128e-17,  -6.23567224e-16]),  
        array([ 0.37431874,  0.18094303,  0.01219871, ...,  0.06501856, 0.04855516,  0.02439321]),  
        array([ 0.41087302,  0.16478851,  0.03396399, ...,  0.09511273, 0.04077713,  0.03945513]),  
        array([ 0.17753915,  0.07025565,  0.09136909, ...,  0.03396507, 0.01379266,  0.01645722]),  
        array([ 0.40605587,  0.05915388,  0.03767763, ...,  0.08981079, 0.05452031,  0.01725399]),  
        array([ 0.        ,  0.        ,  0.        , ...,  0.00579303, 0.02053979,  0.0019091 ]),  
        array([ 0.31550735,  0.11988131,  0.07716529, ...,  0.09815158, 0.03058497,  0.02236517]),  
        array([  0.00000000e+00,   0.00000000e+00,   0.00000000e+00, ..., -3.51175682e-16,   1.31619418e-03,   2.86127901e-16]),  
        array([ 0.21381704,  0.22352378,  0.11568828, ...,  0.06311083, 0.02696666,  0.00402261]),  
        array([ 0.17480064,  0.1469145 ,  0.16336016, ...,  0.05614001, 0.03244093,  0.00524034]),  
        array([ 0.        ,  0.        ,  0.        , ...,  0.03089959, 0.00509584,  0.00247698]),  
        array([ 0.04711166,  0.0218663 ,  0.05316   , ...,  0.04214594, 0.04892439,  0.25840958]),  
        array([ 0.05357464,  0.00530857,  0.07162301, ...,  0.06802692, 0.08331959,  0.26619977])] 

然后我跑了

print img_hogs 

输出是:

[ array([ 0.27621208,  0.11038658,  0.10698133, ...,  0.08661556, 0.04612063,  0.0280782 ]) 
 array([  0.00000000e+00,   0.00000000e+00,   0.00000000e+00, ..., -1.29909838e-15,  -7.01780982e-17,  -1.24900943e-15]) 
 array([ 0.0503603 ,  0.1497235 ,  0.2372957 , ...,  0.07249325, 0.04545541,  0.00903818]) 
 array([ 0.27299191,  0.13122109,  0.0719268 , ...,  0.0848522 , 0.04789403,  0.01387038]) 
 array([  0.00000000e+00,   0.00000000e+00,   0.00000000e+00, ..., 3.32140617e-17,  -6.58924128e-17,  -6.23567224e-16]) 
 array([ 0.37431874,  0.18094303,  0.01219871, ...,  0.06501856, 0.04855516,  0.02439321]) 
 array([ 0.41087302,  0.16478851,  0.03396399, ...,  0.09511273, 0.04077713,  0.03945513]) 
 array([ 0.17753915,  0.07025565,  0.09136909, ...,  0.03396507, 0.01379266,  0.01645722]) 
 array([ 0.40605587,  0.05915388,  0.03767763, ...,  0.08981079, 0.05452031,  0.01725399]) 
 array([ 0.        ,  0.        ,  0.        , ...,  0.00579303, 0.02053979,  0.0019091 ]) 
 array([ 0.31550735,  0.11988131,  0.07716529, ...,  0.09815158, 0.03058497,  0.02236517]) 
 array([  0.00000000e+00,   0.00000000e+00,   0.00000000e+00, ..., -3.51175682e-16,   1.31619418e-03,   2.86127901e-16]) 
 array([ 0.21381704,  0.22352378,  0.11568828, ...,  0.06311083, 0.02696666,  0.00402261]) 
 array([ 0.17480064,  0.1469145 ,  0.16336016, ...,  0.05614001, 0.03244093,  0.00524034]) 
 array([ 0.        ,  0.        ,  0.        , ...,  0.03089959, 0.00509584,  0.00247698]) 
 array([ 0.04711166,  0.0218663 ,  0.05316   , ...,  0.04214594, 0.04892439,  0.25840958]) 
 array([ 0.05357464,  0.00530857,  0.07162301, ...,  0.06802692, 0.08331959,  0.26619977])] 

请您参考如下方法:

尝试:

imgs = [] 
tmp_hogs = np.zeros((17, 256)) 
# 13 of the images are with vehicles, 4 are without 
labels = [1,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0] 
 
i = 0 
for file in out: 
        filepath = "C:\PATH_TO_SAMPLE_IMAGES\\" + file 
        curr_img = color.rgb2gray(io.imread(filepath)) 
        imgs.append(resize(curr_img,(60,40))) 
        fd, hog_image = hog(curr_img, orientations=8, pixels_per_cell=(16, 16), 
                 cells_per_block=(1, 1), visualise=True) 
        tmp_hogs[i,:] = fd 
        i+=1 
 
img_hogs = tmp_hogs