我試圖將我的數據文件,以LIBSVM文件格式,並得到這個錯誤:numpy的陣列錯誤「錯誤('X.shape [0]和y.shape [0]應該是相同的」
ValueError('X.shape[0] and y.shape[0] should be the same got 1L and 3997L instead.',), <traceback object at 0x000000000DC9E608>
我有一個維度的文件特徵和標籤文件,特徵文件(列車數據)發生了錯誤。我使用相同的代碼來處理不同類型的特徵,因爲它是2D的,所以沒有錯誤。幫我
user_train=np.loadtxt(user_train)
train_label=np.loadtxt(train_label)
print('imputing training data ...')
imp=Imputer(missing_values='NaN',strategy='mean',axis=1)
imp.fit(user_train)
user_train=imp.transform(user_train)
print(user_train.shape,train_label.shape)
print('normalizing training data ...')
print(user,i)
normalizer=preprocessing.Normalizer().fit(user_train)
user_train_nor=normalizer.transform(user_train)
print('writing trainiing data')
output=(save_fea+'%s_tag_train_nor.txt'%(user))
svm_rank_format=dump_svmlight_file(user_train_nor,train_label,output,zero_based=False)
數組作爲打印的形狀:((1L,3997L):user_train(地物1D),(3997L):標籤還1D)
看起來你只需要將你的數組從(1,3997)改爲(3997,) – Julien