I have a dataframe grouped with dplyr::group_by and formatted into a table with the gt-package, the output looks like this:
Is there a way to separate this table into multiple or sub-tables using a function of the gt-package, so that every group generated with dplyr::group_by is rendered as separate table (i.e. for printing consecutively)?
Unfortunately, gt_split has no option to do this…
library(tidyverse)
library(gt)
results <- split(iris,~Species) |> imap(\(x,y)gt(x) |> tab_header(title=y))
results[1]
results[2] # etc.
#another form
results_in_one <- gt_group(.list=results)Thank very much.
A good idea the first split, transform to a gt object and then re-group those object.
I just tried it, works fine.
However, I find one downside:
The resulting table have differing widths, that they keep, when re-grouped. Including tab_options(table.width = … ) only fixes this downside as long as … is sufficiently large to cover all resulting table widths. If not, you still get tables of differing width.
You can make them 100% as wide as their html container , by adding to the imap portion
... |> imap(\(x,y)gt(x)|>
tab_options(table.width = pct(100)) |>
tab_header(title=y))Yes, that is what I did, but perhaps did not understand correctly.
Is there a way to pipe the result directly into `gt_group?
I tried in several ways, but always get the error: `data` must be a `gt_tbl` object, not a list.