Klassifizierungstest in Scikit-learn, ValueError: Festlegen eines Array-Elements mit einer Sequenz
Verwendung derTutorial über Adaboost in mehreren KlassenIch versuche, einige Bilder zu klassifizieren, die zwei Klassen haben (aber ich nehme nicht an, dass der Algorithmus nicht funktionieren sollte, wenn das Problem binär ist). Dann werde ich meine Samples um weitere Klassen erweitern.
Mein aktueller Test ist ziemlich klein, insgesamt nur 17 Bilder, 10 zum Training, 7 zum Testen.
Im Moment habe ich zwei Klassen:0: no vehicle, 1: vehicle present Ich habe ganzzahlige Bezeichnungen verwendet, da die Trainingsdaten gemäß dem obigen Link aus ganzzahligen Bezeichnungen bestehen.
Ich habe das bereitgestellte Beispiel nur ein wenig bearbeitet, um meine eigenen Bilddateien einzuschließen, aber ich erhalte eine Fehlermeldung.
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.
Das Folgende ist mein Code, angepasst aus dem Beispiel auf der Website von scikit-learn:
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))
p,l.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()
Ich tippte
print tmp_hogs
und die Ausgabe war wie folgt:
[ 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])]
Dann rannte ich los
print img_hogs
und die Ausgabe war:
[ 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])]