Added an option to scale the columns of Xm and store the scaling factors in an additional factor Z. This makes the task of choosing temporal regularization parameters easier and much more stable.
Allows for a full normal prior for the columns of Fm
Add an initialization option in train() and retrain()
TRMF_es() was removed.It was unclear what this was actually doing.
TRMF_coefficients() was renamed to TRMF_columns.
Updated some of the internal linear algebra for 2-5x speedup
TRMF_summary now prints a weighted R2 for weighted data.
Fixed a bug in the component function which only returned the Xm component.
Fixed a bug in NormalizeMatrix which returns NA's in some cases.
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.