imp903 - 1 year ago 150

Python Question

So, I'm not entirely sure what's going on here, but for whatever reason Python is throwing this at me. For reference, it's part of a small neural network I'm building for fun, but it uses a lot of np.array and such, so there's a lot of matrices being thrown around, so I think it's creating some sort of data type clash. Maybe somebody can help me figure this out, because I've been staring at this error for too long without being able to fix it.

`#cross-entropy error`

#y is a vector of size N and output is an Nx3 array

def CalculateError(self, output, y):

#calculate total error against the vector y for the neurons where output = 1 (the rest are 0)

totalError = 0

for i in range(0,len(y)):

totalError += -np.log(output[i, int(y[i])]) #error is thrown here

#now account for regularizer

totalError+=(self.regLambda/self.inputDim) * (np.sum(np.square(self.W1))+np.sum(np.square(self.W2)))

error=totalError/len(y) #divide ny N

return error

EDIT: Here's the function that returns the output so you know where that came from. y is a vector of length 150 that is taken directly from a text document. at each index of y it contains an index either 1,2, or 3:

`#forward propogation algorithm takes a matrix "X" of size 150 x 3`

def ForProp(self, X):

#signal vector for hidden layer

#tanh activation function

S1 = X.dot(self.W1) + self.b1

Z1 = np.tanh(S1)

#vector for the final output layer

S2 = Z1.dot(self.W2)+ self.b2

#softmax for output layer activation

expScores = np.exp(S2)

output = expScores/(np.sum(expScores, axis=1, keepdims=True))

return output,Z1

Recommended for you: Get network issues from **WhatsUp Gold**. **Not end users.**

Answer Source

Your `output`

variable is not a `N x 4`

matrix, at least not in **python types** sense. It is a **tuple**, which can only be indexed by a single number, and you try to index by tuple (2 numbers with coma in between), which works only for numpy matrices. Print your output, figure out if the problem is just a type (then just convert to np.array) or if you are passing something completely different (then fix whatever is producing `output`

).

Example of what is happening:

```
import numpy as np
output = ((1,2,3,5), (1,2,1,1))
print output[1, 2] # your error
print output[(1, 2)] # your error as well - these are equivalent calls
print output[1][2] # ok
print np.array(output)[1, 2] # ok
print np.array(output)[(1, 2)] # ok
print np.array(output)[1][2] # ok
```

Recommended from our users: **Dynamic Network Monitoring from WhatsUp Gold from IPSwitch**. ** Free Download**