python - 类型错误 : _reshape_dispatcher() missing 1 required positional argument: 'newshape'

Question

我试图将 CNN 模型拟合到某些输入数据 df，其中 df 是 (921 x10170)。

我的代码:

from numpy import *
X = df.iloc[:,0:10165]  
X = X.to_numpy()                    
X = reshape([X.shape[0], X.shape[1], 1])
X_1 = X[:,0:10080,:]          
X_2 = X_train[:,10080:10165,:].reshape(921,85)   
Y = df.iloc[:,10168:10170]

并产生错误:

<小时/>

TypeError Traceback (most recent call last) in () 17 X = df.iloc[:,0:10165] 18 X = X.to_numpy()
---> 19 X = reshape([X.shape[0], X.shape[1], 1]) 20 X_train_1 = X[:,0:10080,:] 21 X_train_2 = X_train[:,10080:10165,:].reshape(921,85)

<array_function internals> in reshape(*args, **kwargs)

TypeError: _reshape_dispatcher() missing 1 required positional argument: 'newshape'

我尝试了几种方法，但都不起作用。这里有什么问题呢？谢谢

Answer 1

您忘记提及您要在这里 reshape 的内容:

X = reshape([X.shape[0], X.shape[1], 1])

您需要将其更改为:

X = X.reshape((X.shape[0], X.shape[1], 1))

它会解决你的问题