Creating a dynamic number of flexdashboard tabs using reactive() - r

Question

I am wondering if it is possible to create a number of tabs in flexdashboard which change according to some user input. As a toy example, I am trying to plot histograms of every numeric numeric column of a user-selected dataset, with each numeric column getting its own tab for a histogram. If the user changes the dataset, the dashboard should update and increase/decrease the number of tabs in the tabset, if the number of numeric columns in the newly selected dataset changes. For simplicity, I only included two datasets.
---
title: "Reactive GGPlot Tabs"
output:
flexdashboard::flex_dashboard:
theme:
version: 4
bootswatch: cosmo
runtime: shiny
---
```{r}
pacman::p_load(tidyverse, shiny, flexdashboard)
```
Sidebar {.sidebar}
-------------------------------------
<h5>Required Input</h5>
```{r}
selectInput(inputId = "dataset", label = "Data:",
choices = c('mtcars', 'iris'), selected = 'iris')
dataset = reactive({
set = as.character(input$dataset)
if(set=='iris'){data = iris} else{data = mtcars}
return(data)
})
```
```{r setup}
chart = reactive({
data = dataset()
num_cols <- unlist(lapply(data, is.numeric))
data = data[,num_cols] # only keep numeric columns for histogram
labels <- data %>% names # these will serve as labels for each tab
chart <- purrr::map(.x = data, ~ggplot(data, aes(x=.x))+geom_histogram()+theme_bw()) %>% setNames(labels) # assign names to each element to use later as tab titles
return(chart)
})
```
## Column {.tabset .tabset-fade}
```{r}
out <- reactive({
lapply(seq_along(chart()), function(i) {
a1 <- knitr::knit_expand(text = sprintf("### %s\n", names(chart())[i])) # tab header, auto extracts names of `chart`, %s indicates string
a2 <- knitr::knit_expand(text = "\n```{r, message = F, warning = F}") # start r chunk
a3 <- knitr::knit_expand(text = sprintf("\nchart()[[%d]]", i)) # extract graphs by "writing" out `chart[[1]]`, `chart[[2]]` etc. to be rendered later, %d indicates integer variable
a4 <- knitr::knit_expand(text = "\n```\n") # end r chunk
paste(a1, a2, a3, a4, collapse = '\n') # collapse together all lines with newline separator
})
})
```
`r reactive({knitr::knit(text = paste(out(), collapse = '\n'))})`
When this code is run, the dashboard displays the markdown chunk code, but does not render plots or tabs.
I am able to sightly modify this code to work without reactive, but I need the number of tabs to increase/decrease according to the chosen dataset. Also, I am not allowed to use facet_wrap.
Is this possible?