Tom Wenseleers

Pull request to support type="link" for nnet::multinom models

26

This adds `type="link"` for `nnet::multinom` models (as is provided by `predict.mblogit`, with logits of reference category normalized to zero & dropped); it also fixes a bug for `type="probs"` (this was...

Automatically split downloads in chunks for queries with >4000 records

1

Just a small possible enhancement, but would it be possible to have the `download `function automatically split the queries in chunks for searches when the length of `list_of_accession_ids` is >4000....

Would it also be possible to support searching by AA substitutions? E.g. I would like to be able to search for E_T11A,NSP3_S403L (picks up more BA.2.75 than searching by lineage)...

Feature request: simultaneous lasso with l1,infinity or l1,lq norm penalization

1

I was wondering if in the future it might also be possible to support l1,linfinity penalization, which is known as the "simultaneous LASSO", see Turlach 2005, https://www.tandfonline.com/doi/pdf/10.1198/004017005000000139 Liu et al...

nnlm function slower than R's nnls or addreg's nnpois + support for sparse matrices?

3

I was doing some tests with the nnlm function hoping it would be faster than the inbuilt nnls R function, but if I try with a 200 x 1000 covariate...

I was wondering if it would be hard to allow NNLM to accept adaptive penalty weights, so one could use it to carry out adaptive LASSO or adaptive ridge (as...

Provide any option for inference?

12

More of a feature request - do you think it might be possible to implement any form of statistical inference in `abess`? E.g. provide confidence intervals on coefficients via bootstrapping?...

Incorrect optimal support size in a multitask learning problem with nonnegativity constraints

15

**Describe the bug** I was testing abess on the multitask learning problem below (deconvolution of a multichannel signal using a known point spread function / blur kernel by regressing the...

glum LASSO fit not finishing when same problem fits within 1 second in glmnet

4

Was just testing glum, but ran into a problem where a glum LASSO fit does not appear to finish (or if it does extremely slowly). The data I simulated was...

On my Windows machine the github versions of `RViennaCL `and `gpuR `installed OK, but when I load them ``` library(RViennaCL) library(gpuR) ``` it doesn't detect my NVIDIA RTX A2000 GPU...