computeCellDistance on large dataset costs too much memory

[wuFYaq]

Firstly thanks for the great tool!

We were running cellchat on cellbin level stereo-seq data (one cell per spot 349,316 cells). The computeCellDistance seems cost too much memory over 1TB with following code, and caused out of memory errors.

spatial.locs = Seurat::GetTissueCoordinates(obj, scale = NULL, cols = c("imagerow", "imagecol"))
spatial.locs = spatial.locs/2
d.spatial <- computeCellDistance(coordinates = spatial.locs, 
             ratio = 1, 
             tol = 10)

Could you please check this error? We can supply test coordinate data.

Much thanks,
F.Y

[sqjin]

@wuFYaq We have now replaced the distance computaiton by a fast method.