heat_map_downsample() keeps at most n cells of each group,
and every cell of smaller groups, so that small clusters stay visible next
to large ones in a heatmap of single cells. The cells are drawn with a
fixed seed, so the same cells are kept each time, and the random numbers
of the session are left as they were. heat_map(..., group_max = n)
downsamples the groups of group_by in this way.
Usage
heat_map_downsample(groups, n = 500, seed = 2357)
Arguments
- groups
a vector with the group of each cell (e.g. its cluster).
Cells without a group (NA) are left out.
- n
the largest number of cells to keep from each group, set to 500 by
default.
- seed
the random seed for the draw, set to 2357 by default.
Value
the indices of the cells kept, in increasing order.
Author
Dillon Hammill (dillon.hammill21@gmail.com)
Examples
groups <- rep(c("T cells", "B cells", "NK cells"), c(2000, 300, 40))
keep <- heat_map_downsample(groups, n = 100)
table(groups[keep])
#>
#> B cells NK cells T cells
#> 100 40 100