Make a list of arguments for a function from a character vector-CodePudding

I am making a ComplexHeatmap and I want to show multiple row annotations (i.e., multiple named arguments inside the rowAnnotation function) corresponding to columns of information from my original data. What I am doing now, having 7 columns of information:

library(ComplexHeatmap)


# sample dataframe
EF_data <- data.frame(Package = sample(1:10, 10),
                      Information = sample(1:10, 10),
                      Atoms = sample(1:10, 10),
                      Measure = sample(1:10, 10),
                      Filter = sample(1:10, 10),
                      Analysis = sample(1:10, 10),
                      Averaging = sample(1:10, 10))


# what i'm doing now
left_an_gp <- gpar(fontsize = 8)

left_an = rowAnnotation(Package = anno_text(EF_data$Package, gp = left_an_gp),
                        Information = anno_text(EF_data$Information, gp = left_an_gp),
                        Atoms = anno_text(EF_data$Atoms, gp = left_an_gp),
                        Measure = anno_text(EF_data$Measure, gp = left_an_gp),
                        Filter = anno_text(EF_data$Filter, gp = left_an_gp),
                        Analysis = anno_text(EF_data$Analysis, gp = left_an_gp),
                        Averaging = anno_text(EF_data$Averaging, gp = left_an_gp),
                        gap = unit(0.5, "mm"))

# and then draw the heatmap with the data from a matrix etc

What I want to do is to generate the rowAnnotation using, for example, the columns(EF_data) character vector. Every argument inside rowAnnotate is the same except for its name and the name of the column after the $, so I wonder if there is a more compressed way of doing this. I have tried to search for one but I guess I'm not using the right terminology to find an approach.

I imagine it would be something like:

apply(
  eval(
    "{} = anno_text(EF_data${}, gp = left_an_gp)"
    ), 
  columns(EF_data)
)

but I don't quite know how to even start to try to do it correctly.

CodePudding user response：

You don't need to resort to parsing and evaluating text. Instead, you can build the arguments to the call using lapply and invoke the function with do.call

args <- lapply(EF_data, function(x) row_anno_text(x, gp = left_an_gp))
args <- setNames(args, names(EF_data))
args$gap <- unit(0.5, "mm")
left_an <- do.call(rowAnnotation, args)

left_an
#> A HeatmapAnnotation object with 7 annotations
#>   name: heatmap_annotation_0 
#>   position: row 
#>   items: 10 
#>   width: 24.9681777777778mm 
#>   height: 1npc 
#>   this object is subsettable
#> 
#>         name annotation_type color_mapping              width
#>      Package     anno_text()               3.13831111111111mm
#>  Information     anno_text()               3.13831111111111mm
#>        Atoms     anno_text()               3.13831111111111mm
#>      Measure     anno_text()               3.13831111111111mm
#>       Filter     anno_text()               3.13831111111111mm
#>     Analysis     anno_text()               3.13831111111111mm
#>    Averaging     anno_text()               3.13831111111111mm

^{Created on 2022-05-09 by the reprex package (v2.0.1)}