I’ m doing a filter function:

```
def smoothfilter(extremeBands, CHSC, X):
filteredSmooth = torch.zeros(np.size(CHSC, 1) * np.size(CHSC, 2), X.shape[3],
dtype=torch.float64)
for i in range(np.size(CHSC, 1) * np.size(CHSC, 2)-1):
temp = extremeBands[i, :]
for j in range(X.shape[3]):
if j == 1:
temp[j] = torch.mean(extremeBands[i, j - 1:j + 2])
if j == (X.shape[3] - 2):
temp[j] = torch.mean(extremeBands[i, j - 1:])
if j > 1 and j != (X.shape[3] - 3):
temp[j] = torch.mean(extremeBands[i, j - 2:j + 3])
if j == (X.shape[3] - 3):
temp[j] = torch.mean(extremeBands[i, j - 2:])
filteredSmooth = temp
return filteredSmooth
```

I get the IndexError in the line: temp[j] = torch.mean(extremeBands[i, j - 2:j + 3]). Does someone know how to fix it?