Bug in lbm_bifunc

[zyh722]

Hi, when I run the lbm_bifunc code listed below:

lbm_res = lbm_bifunc(data1, K=4, L=3, display=False, nbasis = 5,
                     basis_name = 'spline', init='kmeans')

I encounter a bug:

ValueError: could not broadcast input array from shape (40,5) into shape (5,)

How to solve this?

[Barry57]

Thanks for your attention! This is because, in this case, splines performs very poorly for classification; within this method we recommend using Fourier series instead. Splines often cluster one class into a single observation, so their use is discouraged.

[zyh722]

Ok, I get it. Thanks!