make_heatmap_for_table: Make a heatmap with one column for each cluster in 'unique(...
In maehrlab/thymusatlastools: Tools for analysis of single-cell transcriptomic data

Make a heatmap with one column for each cluster in 'unique( Seurat::FetchData(dge, ident.use)[[1]])' and one row for every gene in 'genes_in_order'.

make_heatmap_for_table(dge, genes_in_order, desired_cluster_order = NULL,
  ident.use = "ident", labels = NULL, aggregator = mean,
  normalize = "row", norm_fun = div_by_max, genes_to_label = NULL,
  main = "Genes aggregated by cluster", return_type = "plot")

`dge`	Seurat object
`desired_cluster_order`	Levels of FetchData(dge, ident.use)[[1]], ordered how you want them to appear.
`ident.use`	Variable to aggregate by.
`labels`	"regular" (all labels), "stagger" (all labels, alternating left-right to fit more genes), or "none".
`aggregator`	Function to aggregate expression within a group of cells. Try mean or prop_nz.
`normalize`	"row", "column", "both", or "none". Normalization function gets applied across the axis this specifies.
`norm_fun`	Function to use for normalization. Try div_by_max or standardize.
`genes_to_label`	A (small) subset of genes to label. If set, nullifies labels arg.
`main`	Title of plot.
`return_type`	"plot" or "table". If "table", then instead of returning a heatmap, this returns the underlying matrix of normalized, cluster-aggregated values. If anything else, returns a ggplot. If the cluster's expression values are stored in 'x', then 'aggregator(x)' gets (normalized and) plotted. Optional parameter 'desired_cluster_order' gets coerced to character. It should be a subset of 'unique(Seurat::FetchData(dge, ident.use))' (no repeats).