search

peterwittek / somoclu

Massively parallel self-organizing maps: accelerate training on multicore CPUs, GPUs, and clusters
https://peterwittek.github.io/somoclu/
MIT License
268 stars 70 forks source link

Dimensions .. issue #114

Closed Tortoise17 closed 6 years ago

Tortoise17 commented 6 years ago

Dear Respected Friends.

Good Morning. I want to ask a question. After the clustering in SOMOCLU, when I get the get_surface_state to get BMUS.. , i get the array is of like 1523502 and the input was (951 by 801 = 761751). and the surface should have the values 761751. May I ask where is my mistake and why the values are double. How to resolve the issue? I mean one column is neuron number and other is the BMUS. so, how to get it plotted without neuron number?? my workflow is

ax = som.get_surface_state() a = som.get_bmus(ax) test = np.array[(a)] em = np.arange(test).reshape((951, 801)))

and error is

File "", line 1, in em = np.arange(test).reshape((951, 801))

TypeError: only length-1 arrays can be converted to Python scalars .. This is because the file has neurons and BMUS values. how to get it displayed. if you can guide ..

Basically, I want to see the BMUS in the original space back with dimensions. How to see them as map view. This is problem for me. ++++++++++++++++++++++++++++++++++++++++++++

Please help

Looking forward.

Waiting for help.

Tortoise17 commented 6 years ago

OK.. for now I think I realized what mistake I have done. Thank you for help. if something comes up. I will ask again..

Very nice tool.

Tortoise17 commented 6 years ago

At the moment. I have realized my mistake. Thank you..