do_BarPlot() | Categorical proportions

Bar plots display counts or proportions of categorical variables. Use them to compare cell type composition across samples, conditions, or any grouping variable.

Basic usage

Count cells per cluster:

p <- SCpubr::do_BarPlot(sample = sample,
                        group.by = "Annotation")
p

Split by a second variable

Show how clusters are distributed across samples:

p <- SCpubr::do_BarPlot(sample = sample,
                        group.by = "Annotation",
                        split.by = "ID",
                        position = "stack")
p

Think of it as: - group.by: What creates the bars (total counts per value) - split.by: How each bar is subdivided

Show proportions

Use position = "fill" to normalize bars to the same height:

p <- SCpubr::do_BarPlot(sample = sample,
                        group.by = "Annotation",
                        split.by = "ID",
                        position = "fill",
                        flip = TRUE)
p

Add total counts

Display the total cell count on each bar:

p <- SCpubr::do_BarPlot(sample = sample,
                        group.by = "Annotation",
                        split.by = "ID",
                        position = "fill",
                        flip = TRUE,
                        add.n = TRUE,
                        add.n.size = 4,
                        add.n.expand = c(0, 1.15))
p

Order bars

Order by total count or by a specific group’s value:

# Order by total count
p <- SCpubr::do_BarPlot(sample = sample,
                        group.by = "Annotation",
                        order = TRUE)

p

Facet by a third variable

Break down bars by an additional grouping:

p <- SCpubr::do_BarPlot(sample = sample,
                        group.by = "Annotation",
                        split.by = "ID",
                        facet.by = "subtype",
                        position = "fill",
                        flip = TRUE)
p

Return data

Get the underlying count/proportion data:

result <- SCpubr::do_BarPlot(sample = sample,
                             group.by = "Annotation",
                             split.by = "ID",
                             return_data = TRUE)

head(result$Data)
#> # A tibble: 6 × 4
#>   ID             n    freq Annotation      
#>   <chr>      <dbl>   <dbl> <chr>           
#> 1 ATRT-MYC-1     8 0.00756 CP-like         
#> 2 ATRT-MYC-1     0 0       Cilia-like      
#> 3 ATRT-MYC-1    69 0.0652  Hypoxic         
#> 4 ATRT-MYC-1    71 0.0671  IPC-like        
#> 5 ATRT-MYC-1     0 0       Immune-like     
#> 6 ATRT-MYC-1    32 0.0302  Mesenchymal-like

Parameter reference

Note

For parameters shared across many functions (color palettes, typography, legend styling), see Shared features.

Core parameters

Parameter Description Default
group.by Metadata column for bars —
split.by Metadata column to subdivide bars NULL
facet.by Metadata column to facet by NULL
position "stack" (counts) or "fill" (proportions) "stack"

Ordering

Parameter Description Default
order Order bars by descending count FALSE
order.by Value of split.by to order by NULL

Count labels

Parameter Description Default
add.n Show total counts on bars FALSE
add.n.size Label text size 4
add.n.face Label font face "bold"
add.n.expand Y-axis expansion for labels c(0, 1.15)

Output

Parameter Description Default
return_data Return data instead of plot FALSE

See also