Minor suggestion on spatialFeaturePlot()

[mvfki]

I'm working on a Visium HD dataset, which is relatively large and has 85k cells. And when I wanted to see some gene expression with spatialFeaturePlot(), it immediately warned me saying

Warning message:
In asMethod(object) :
  sparse->dense coercion: allocating vector of size 12.1 GiB

I dove into the source code and saw data <- as.matrix(object@data) in the first line of the function body. I have a sparse matrix presented when constructing the CellChat object, and converting the whole thing into a dense form is obviously the cause.

My personal suggestion for a better implementation would be first to check the features input and only subset the expression of specified ones, which are the only ones necessary to be converted to dense.

[sqjin]

@mvfki Thanks for the suggestion. This can be modified as data <- object@data. We will update it.