SydneyBioX / scMerge

Statistical approach for removing unwanted variation from multiple single-cell datasets
https://sydneybiox.github.io/scMerge/
66 stars 13 forks source link

Accepting other matrix classes (e.g. SparseMatrix/DelayedArray) #9

Closed kevinwang09 closed 5 years ago

kevinwang09 commented 5 years ago

In scMerge V1.1.0, we only accept SingleCellExperiment objects with each slot being class matrix. The package does not work if class SparseMatrix is used, nor DelayedArray.

https://bioconductor.org/packages/release/bioc/vignettes/DelayedArray/inst/doc/02-Implementing_a_backend.html

This feature will be beneficial for large SingleCellData.

kevinwang09 commented 5 years ago

Also, add another function that removes rows and columns of pure zeroes for the convenience of users.

kevinwang09 commented 5 years ago

The delayedarray branch partially solves this problem. The first Travis passed commit is SydneyBioX@09f5b73, though this is not totally robust

kevinwang09 commented 5 years ago

The sparsearray branch partially solves this problem. The first Travis passed commit is  SydneyBioX@249a4cccfdb11234b045d721f67dde25da4511cf, though this is not totally robust.