kimtonyhyun / analysis

2 stars 0 forks source link

Kimth/minor tweaks #243

Closed kimtonyhyun closed 7 years ago

kimtonyhyun commented 7 years ago

@ahwillia fit_cpd uses non-negative ALS by default -- which seems to me appropriate for the prefrontal experiment.

I also modified the tensor doc to factor out cpd=models(1,15).decomp for readability.

Finally, visualize_neuron_ktensor uses "k" to index factors than "r" in the plot. I propose to reserve "r" for the rank of the CPD model.

@forea You can look at docs/tensor.md to get an introduction to the tensor factorization code.

kimtonyhyun commented 7 years ago

@ahwillia Changed back to "r".