plot_hexbin_meta: Plot of meta data of single cell data in bivariate hexagon...
In schex: Hexbin plots for single cell omics data

Description Usage Arguments Details Value Examples

Plot of meta data of single cell data in bivariate hexagon cells.

plot_hexbin_meta(
  sce,
  col,
  action,
  no = 1,
  colors = NULL,
  title = NULL,
  xlab = NULL,
  ylab = NULL,
  na.rm = FALSE
)

`sce`	A `SingleCellExperiment` or `Seurat-class` object.
`col`	A string referring to the name of one column in the meta data of sce by which to colour the hexagons.
`action`	A string specifying how meta data of observations in binned hexagon cells are to be summarized. Possible actions are `majority`, `prop`, `prop_0`, `mode`, `mean` and `median` (see details).
`no`	An integer specifying which level to plot of the column. Only in effect when `action=prop`.
`colors`	A vector of strings specifying which colors to use for plotting the different levels in the selected column of the meta data. Only in effect when the selected `action="majority"`.
`title`	A string containing the title of the plot.
`xlab`	A string containing the title of the x axis.
`ylab`	A string containing the title of the y axis.
`na.rm`	Logical indicating whether NA values should be removed.

This function plots any column of the meta data in the hexagon cell representation calculated with make_hexbin. The chosen meta data column is summarized by one of six actions majority, prop, prop_0, mode, mean and median:

majority: Returns the value of the majority of observations in the bin. The associated meta data column needs to be a factor or character.
prop: Returns the proportion of each level or unique character in the bin. The associated meta data column needs to be a factor or character.
prop_0: Returns the proportion of observations in the bin greater than 0. The associated meta data column needs to be numeric.
mode: Returns the mode of the observations in the bin. The associated meta data column needs to be numeric.
mean: Returns the mean of the observations in the bin. The associated meta data column needs to be numeric.
median: Returns the median of the observations in the bin. The associated meta data column needs to be numeric.

A ggplot2{ggplot} object.

#' # For Seurat object
library(Seurat)
data("pbmc_small")
pbmc_small <- make_hexbin(pbmc_small, 10, dimension_reduction = "PCA")
plot_hexbin_meta(pbmc_small, col="RNA_snn_res.1", action="prop", no=1)
# For SingleCellExperiment object
## Not run: 
library(TENxPBMCData)
library(scater)
tenx_pbmc3k <- TENxPBMCData(dataset = "pbmc3k")
rm_ind <- calculateAverage(tenx_pbmc3k) < 0.1
tenx_pbmc3k <- tenx_pbmc3k[-rm_ind, ]
colData(tenx_pbmc3k) <- cbind(
  colData(tenx_pbmc3k),
  perCellQCMetrics(tenx_pbmc3k)
)
tenx_pbmc3k <- logNormCounts(tenx_pbmc3k)
tenx_pbmc3k <- runPCA(tenx_pbmc3k)
tenx_pbmc3k <- make_hexbin(tenx_pbmc3k, 20, dimension_reduction = "PCA")
plot_hexbin_meta(tenx_pbmc3k, col = "total", action = "median")

## End(Not run)