I would consider that a missing feature, even if it looks like a basic one.
The support for sparse arrays has not been a priority in the development of Theano for a while, unfortunately. On Mon, Nov 07, 2016, fstab wrote: > Hi, > > I have been running a few experiments with sparse matrices and I was > wondering: why is comparison of a sparse matrix with a scalar not allowed? > > In other words, when I do `S > 0.5` , with S being a theano CSR matrix, an > AssertionError from sparse/basic.py is raised: assert y.type.ndim == 2 > > Why would this be a requirement? Is it a feature, work in progress, or a > bug? > > looking forward to your reply. > > regards, > -Francesco > > -- > > --- > You received this message because you are subscribed to the Google Groups > "theano-users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. -- Pascal -- --- You received this message because you are subscribed to the Google Groups "theano-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
