Immanuel Abdi

Usage of sparse attention within GENA_LM

2

Hi Yuri and Veniamin. Thanks for sharing your work, I very much enjoyed reading about how you were able to extend context length much further than what can be typically...

Macs peak calling

3

# Description Add macs2/macs3 peak calling Utilize a multi-threaded approach separated by cluster within `prep_macs_inputs()`. Allow users to choose a specific step to run (ie "prep-inputs", "run-macs", "read-outputs") to allow...

[r][cpp] add pseudobulking to matrices

4

# Description Added function `matrix_quantile_per_cell()` to find the nth quantile value of each cell in a matrix. Allows for clipping using `min_by_row()` and `min_by_col()` Added `pseudobulk_matrix()` with option to clip...

Ia/insertion radix sorting

1

# Description Previously we added a change for `InsertionIterator` to use a priority queue, as we were experiencing memory errors with the original vector + radixsort method. However, the priority...

[cpp] Fix matrix Windows/linux matrix compatibility issues

1

Writing matrices on a Windows setup previously were unable to be read on Linux (and probably mac) setups, with the following error `Error in eval(exprs, envir): Matrix has unrecognized version...

Normalizations, Feature Selections, DimReduction S3 class, (Iterative) Lsi Implementation, Partial function framework, Clustering changes and Framework

1

Meta-PR that describes the culmination of efforts of multiple features. Included and previously described features: - Iterative LSI design docs (#167) - Log, tf-idf normalization functions (#168) - Feature selection...

This PR provides function examples for most of the functions in the `Single Cell plots` and `Genomic track plots` section. This PR also provides a function (`render_plot_from_storage()`) that allows for...

Add examples for every `IterableMatrix` method together in the `IterableMatrix-methods.Rd`. Using a super simple example of transforms on `as(matrix(1:25, nrow = 5), "IterableMatrix")`. One thing that I'm thinking is it...

Note, there is still a bug in `tile_matrix()` that hasn't been fixed when you have filtered some chromosomes out. Caused me to use `subset = FALSE` for them.