I am trying to cluster similar data points using quicksom package which is GPU based SOM using Pytorch. All the functions are running smoothly except when I am using som.predict() function that calculates quantization errors and topographic errors. But the function stops in the middle and throws the following errors:
When I am using batch size =25
‘IndexError: too many indices for tensor of dimension 1’
And when I am using default batch size = 100
‘ValueError: too many values to unpack (expected 2)’
I have attached the snippet of my data, and the errors I am getting.