I’m trying to sample from a one-hot encoded matrix that’s got a shape of roughly 225,000 x 1,000. Using the GLM function within pymc3 my goal is to extract the distributions for each of the coefficients from the matrix, while also incorporating gaussain priors for each of them as well. I’ve included…

Sparse array glm fit in pymc3 very long with NUTS

junpenglao October 27, 2020, 7:46am 7

Could you try casting matrix into theano? Seems there is a memory error when theano is trying to do that internally.

Topic		Replies	Views
Possible Performance Issue Using Sparse Matrices in Bipartite Noisy-Or Model Questions from_github , theano	10	2631	September 15, 2017
Sparse Gradient Questions development	4	799	February 12, 2022
Memory issues with creating simple regression model Questions	4	1959	June 17, 2019
Larger dataset linear regression GPU memory problems with NUTS for larger N and D (batch training?) Questions theano	2	1868	October 3, 2017
Theano sparse throws error on sampling Questions theano	2	593	April 20, 2018