Dynamically generating figures with captions in Word output from Rmarkdown document - r-markdown

I'm trying to generate a Word document with figures. I use {officedown} and {officer} packages. I need to do it dynamically, in a loop since I don't know how many figures there will be. This is my Rmarkdown code:
---
output:
officedown::rdocx_document:
plots:
caption:
style: Table Caption
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE, message = FALSE, warning = FALSE)
```
```{r}
library(dplyr)
library(flextable)
library(officer)
```
```{r}
block_toc(seq_id = "fig")
```
```{r results='asis'}
ttp <- c(3, 7)
test_items <- c("item A", "item B", "item C")
fpars <- lapply(test_items, function(ti) {
fpar(
ftext("Application of "),
ftext(ti),
ftext(" Variable text - describe any test item-related effects"),
ftext(" (see "),
run_reference("fig:results1"),
ftext(")."),
fp_p = fp_par(padding.bottom = 12)
)
})
do.call(block_list, fpars)
```
```{r}
titles <- lapply(seq_len(length(ttp)), function(i) {
sprintf(
"My custom figure caption with %s, having %s side effects",
paste(test_items, collapse = ", "),
ttp[i]
)
})
```
```{r}
tmps <- lapply(seq_len(length(ttp)), function(i) {
tmp <- tempfile(fileext = ".png")
png(tmp, width = 6, height = 5, units = "in", res = 120)
plot(iris[sample(1:150, 30), i + 1:2])
dev.off()
return(tmp)
})
```
```{r}
fpars <- lapply(seq_len(length(ttp)), function(i) {
fpar(
run_autonum(
seq_id = "fig",
pre_label = "Figure ",
bkm = paste0("fig:results", i),
bkm_all = TRUE,
prop = fp_text(bold = TRUE, font.size = 12)
),
titles[[i]],
external_img(src = tmps[[i]], width = 6, height = 5)
)
})
do.call(block_list, fpars)
```
The problem is when I generate the table of figures in the rendered document. It looks like this:
An entry is kept together with the image itself, I don't know why.
I save temporary png files to be able to use them inside fpar function. Using plot function directly inside fpar causes bad effects. Maybe there's another/better way?
I found this construction useful, but unfortunately it puts captions under the figures by default. My goal is figure captions behave more like table captions, i.e. a caption is above a figure.
```{r fig.cap=unlist(titles)}
plot(iris[1:10, 1:2])
```
How can I generate the plots with captions dynamically with {officedown} or {officer} packages?

The only issue is about the last lapply, you need to put the image in a separate paragraph, so a for loop will be easier to stack 2 paragraphs into a list instead of 1 for each iteration.
---
output:
officedown::rdocx_document:
plots:
caption:
style: Table Caption
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE, message = FALSE, warning = FALSE)
```
```{r}
library(dplyr)
library(flextable)
library(officer)
library(officedown)
```
```{r}
block_toc(seq_id = "fig")
```
```{r}
ttp <- c(3, 7)
test_items <- c("item A", "item B", "item C")
fpars <- lapply(test_items, function(ti) {
fpar(
ftext("Application of "),
ftext(ti),
ftext(" Variable text - describe any test item-related effects"),
ftext(" (see "),
run_reference("fig:results1"),
ftext(")."),
fp_p = fp_par(padding.bottom = 12)
)
})
do.call(block_list, fpars)
```
```{r}
titles <- lapply(seq_len(length(ttp)), function(i) {
sprintf(
"My custom figure caption with %s, having %s side effects",
paste(test_items, collapse = ", "),
ttp[i]
)
})
```
```{r}
tmps <- lapply(seq_len(length(ttp)), function(i) {
tmp <- tempfile(fileext = ".png")
png(tmp, width = 6, height = 5, units = "in", res = 120)
plot(iris[sample(1:150, 30), i + 1:2])
dev.off()
return(tmp)
})
```
```{r}
fpars <- list()
for (i in seq_along(ttp)) {
fpars[[length(fpars)+1]] <- fpar(
run_autonum(
seq_id = "fig",
pre_label = "Figure ",
bkm = paste0("fig:results", i),
bkm_all = TRUE,
prop = fp_text(bold = TRUE, font.size = 12)
),
titles[[i]]
)
fpars[[length(fpars)+1]] <- fpar(
external_img(src = tmps[[i]], width = 6, height = 5)
)
}
do.call(block_list, fpars)
```
This is an answer for "how to do it with officer" from scratch.
But the following is much more simple. There is an option fig.topcaption that do the same thing.
---
output:
officedown::rdocx_document:
plots:
caption:
style: Table Caption
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE, message = FALSE, warning = FALSE)
```
```{r}
library(dplyr)
library(flextable)
library(officer)
library(officedown)
```
```{r}
block_toc(seq_id = "fig")
```
```{r fig.cap = "miaou", fig.topcaption=TRUE}
plot(cars)
```
```{r fig.cap = "ouaf", fig.topcaption=TRUE}
plot(cars)
```

Related

using a for loop to generate a series of plots in new powerpoint slides using quarto

I am creating a quarto powerpoint presentation and want to use a dataset to iterate over all the classes in a variable and do a plot for each class on a new slide. However, my plot doesn't seem to show up
---
title: "test_quarto_presentation"
format: pptx
editor: visual
---
## Quarto
Trying to iterate over each species and print the distributions on slides
```{r}
library(tidyverse)
```
```{r, results='asis'}
for (species in unique(iris$Species)){
cat(paste0('## ', species))
print("test")
print(iris %>%
filter(Species == "virginica") %>%
ggplot(aes(x = Petal.Length)) +
geom_histogram())}
```
Here is an option by producing the plots beforehand using purrr based on this discussion:
---
title: "test_quarto_presentation"
format: pptx
---
```{r, include=FALSE}
library(tidyverse)
spec_name <- unique(iris$Species)
make_hist <- function(spec) {
iris |>
filter(Species == spec) |>
ggplot(aes(x = Petal.Length)) +
geom_histogram()
}
list_hist <- map(spec_name, make_hist)
df <- tibble(spec = spec_name, plots = list_hist)
```
```{r}
#| output: asis
res <- pmap_chr(df, \(spec, plots) {
knitr::knit_child(text = c(
"## `r spec`",
"```{r}",
"#| echo: false",
"plots",
"```",
""), envir = environment(), quiet = TRUE)
})
cat(res, sep = '\n')
```
Output:

Flexdashboard and absolute panel on a lealfet

Is there a way to include an absolute panel like in this example (https://shiny.rstudio.com/gallery/superzip-example.html) to a flexdashboard (on a leaflet) ?
The idea would be to have a mobile panel dedicated to the leaflet output instead of a sidebar panel.
The absolute panel example here based on a shiny example (with ui and server parts)
library(shiny)
ui <- shinyUI(bootstrapPage(
absolutePanel(
id = "controls", class = "panel panel-default", fixed = TRUE,
draggable = TRUE, top = 60, left = "auto", right = 20, bottom = "auto",
width = 330, height = "auto",
HTML('<button data-toggle="collapse" data-target="#demo">Collapsible</button>'),
tags$div(id = 'demo',
checkboxInput('input_draw_point', 'Draw point', FALSE ),
verbatimTextOutput('summary')))
))
server <- shinyServer(function(input, output, session) {
output$summary <- renderPrint(print(cars))
})
shinyApp(ui = ui, server = server)
a code example below for the Flexdashboard part :
---
title: "Waste Lands - America's forgotten nuclear legacy"
author: Philipp Ottolinger
output:
flexdashboard::flex_dashboard:
theme: journal
social: menu
source_code: embed
---
```{r setup, include = FALSE}
library(flexdashboard)
library(shiny)
library(jsonlite)
library(maptools)
library(ggplot2)
library(tidyr)
library(dplyr)
library(purrr)
library(leaflet)
library(plotly)
sites <- fromJSON(flatten=TRUE,
"https://raw.githubusercontent.com/ottlngr/2016-15/ottlngr/ottlngr/sites.json")
sites$locations <- map(sites$locations, function(x) {
if (nrow(x) == 0) {
data_frame(latitude=NA, longitude=NA, postal_code=NA, name=NA, street_address=NA)
} else {
x
}
})
sites <- unnest(sites)
sites <- sites[complete.cases(sites[,c("longitude", "latitude")]),]
sites$ratingcol <- ifelse(sites$site.rating == 0, "orange",
ifelse(sites$site.rating == 1, "green",
ifelse(sites$site.rating == 2, "red", "black")))
sites$ratingf <- factor(sites$site.rating,
levels=c(3:0),
labels=c("Remote or no potential for radioactive contamination.",
"No authority to clean up or status unclear.",
"Cleanup declared complete.",
"Cleanup in progress."))
sites$campus <- ifelse(grepl("University", sites$site.name) |
grepl("University", pattern = sites$street_address) |
grepl("Campus", sites$street_address), 1, 0)
sites$campuscol <- ifelse(sites$campus == 1, "red", "black")
```
Column {data-width=650}
-----------------------------------------------------------------------
### All sites and their current status
```{r}
leaflet() %>%
addTiles() %>%
fitBounds(-127.44,24.05,-65.30,50.35) %>%
addCircleMarkers(sites$longitude,
sites$latitude,
color = sites$ratingcol,
radius = 6,
fill = T,
fillOpacity = 0.2,
opacity = 0.6,
popup = paste(sites$site.city,
sites$site.name,
sep = "")) %>%
addLegend("bottomleft",
colors = c("orange","green", "red", "black"),
labels = c("Cleanup in progress.",
"Cleanup complete.",
"Status unclear.",
"No potential for radioactive contamination."),
opacity = 0.8)
```
Column {data-width=350}
-----------------------------------------------------------------------
### Number of sites
```{r}
sites %>%
count(ratingf) %>%
plot_ly(type = "bar",
x = ratingf,
y = n,
color = ratingf,
text = paste(n,ratingf,sep=""),
hoverinfo = "text") %>%
layout(xaxis = list(showline = F,
showticklabels = F,
fixedrange = T,
title = ""),
yaxis = list(fixedrange = T,
title = ""))
```
### Sites on campus
```{r}
leaflet() %>%
addTiles() %>%
fitBounds(-127.44,24.05,-65.30,50.35) %>%
addCircleMarkers(sites[sites$campus == 1, ]$longitude,
sites[sites$campus == 1, ]$latitude,
color = sites[sites$campus == 1, ]$campuscol,
radius = 6,
fill = T,
fillOpacity = 0.2,
opacity = 0.6,
popup = paste(sites[sites$campus == 1, ]$site.city,
sites[sites$campus == 1, ]$site.name,
sep = ""))
```
Thanks
Try this.
---
title: "haha"
output:
flexdashboard::flex_dashboard:
orientation: columns
vertical_layout: fill
editor_options:
chunk_output_type: console
runtime: shiny
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
library(flexdashboard)
library(shiny)
library(leaflet)
```
# without container-fluid
### Sites on campus
```{r}
df <- data.frame(NY = c(-74.418997, 43.257408), CA = c(-120.765285, 35.604380))
renderLeaflet(mapfunction())
```
```{r}
absolutePanel(
draggable = TRUE, top = "15%", left = "auto", right = "5%", bottom = "auto",
width = '30%', height = 'auto',
style = "background: orange; opacity: 0.9",
p(strong("some text")),
selectInput("someinput", label = "location", choices = c("NY", "CA"))
)
```
### server
```{r}
mapfunction <- reactive({
leaflet() %>%
addTiles() %>%
fitBounds(-127.44,24.05,-65.30,50.35) %>%
addMarkers(lng = df[[input$someinput]][1], lat = df[[input$someinput]][2])
})
```
# with container-fluid
```{r}
shinyApp(
fluidPage(
leafletOutput(outputId = "somemap"),
absolutePanel(
draggable = TRUE, top = "15%", left = "auto", right = "5%", bottom = "auto",
width = '30%', height = 'auto', fixed = TRUE,
style = "background: orange; opacity: 0.9",
p(strong("some text")),
selectInput("someinput", label = "location", choices = c("NY", "CA"))
)
),
server = function(input, output, session){
df <- data.frame(NY = c(-74.418997, 43.257408), CA = c(-120.765285, 35.604380))
output$somemap <- renderLeaflet({
leaflet() %>%
addTiles() %>%
fitBounds(-127.44,24.05,-65.30,50.35) %>%
addMarkers(lng = df[[input$someinput]][1], lat = df[[input$someinput]][2])
})
}
)
```
If you need to use interactive components from shiny, like XXinput, you need to specify runtime: shiny on the top, otherwise, you can delete this line.
I use reactive as the simplest server part. If you want to use a more complex server (logic), e.g. several components interact together, you need to write the actual server function. I would suggest just write a shiny app instead of a flexdashboard.
unfortunately, components in flexdash are not inside container-fluid class which with this can allow you to drag the panel. There may be a way to work around, you can search for it. Look at the last chunck, I inserted a actual shiny app and the panel is draggable. You should see two tabs when you run the doc, watch the difference. So, if you really want to drag this panel, you should write a "real" shiny app.

How do you filter a data frame in a shiny document and display a datatable?

I'm trying to filter a data frame and then do some simple ggplots off of the data. I've tried to leverage the R studio example on Shiny documents along with the following SO post on the subject:
Reactively filtering/subsetting a data frame in shiny
Here is my code.
---
title: "Shiny Filter Test"
author: "Novice"
date: "12/13/2019"
output: html_document
runtime: shiny
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```
```{r}
library(tidyverse)
library(shiny)
inputPanel(
selectInput("n_break", label = "Number of bins:",
choices = c(10, 20, 35, 50), selected = 10)
)
cdat <- reactive({
data <- tibble(x = c(10,20,35), y = c("a","b","c"))
data %>%
filter(x %in% input$n_break)
output$table <- DT::renderDT({
cdat()
}, options = list(scrollX = TRUE))
})
```
Can anyone point out where I'm going wrong? When I run the code I get my dropdown box, but that is all. No errors. Just no filtered datatable.
Thanks.
The closing brackets of your reactive are at the wrong place. They should close once you have filtered the data.
---
title: "Shiny Filter Test"
author: "Novice"
date: "12/13/2019"
output: html_document
runtime: shiny
---
```{r setup}
knitr::opts_chunk$set(
echo = FALSE
)
```
```{r}
library(tidyverse)
library(shiny)
inputPanel(
selectInput("n_break", label = "Number of bins:",
choices = c(10, 20, 35, 50), selected = 10)
)
cdat <- reactive({
data <- tibble(x = c(10,20,35), y = c("a","b","c"))
data %>% filter(x %in% input$n_break)
})
DT::renderDT({
cdat()
}, options = list(scrollX = is ))
```
A remark on the reactive: if you plan to extend this futher, such that the filtered data is used elsewhere, it makes sense to do the filtering in a reactive function. However, if this is not the case I would just do the filtering inside the renderDT:
---
title: "Shiny Filter Test"
author: "Novice"
date: "12/13/2019"
output: html_document
runtime: shiny
---
```{r setup}
knitr::opts_chunk$set(
echo = FALSE
)
```
```{r}
library(tidyverse)
library(shiny)
data <- tibble(x = c(10,20,35), y = c("a","b","c"))
inputPanel(
selectInput("n_break", label = "Number of bins:",
choices = c(10, 20, 35, 50), selected = 10)
)
DT::renderDT({
data %>% filter(x %in% input$n_break)
}, options = list(scrollX = TRUE))
```

Create slide with render dynamically in Rmarkdown shiny presentation

I want to produce slide dynamically in my Rmarkdown presentation with shiny
Here below the code works.
But as soon as I uncomment the for loop it doesn't.
---
title: "dynamic_plots"
output: ioslides_presentation
runtime: shiny
---
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE)
```
```{r, echo = FALSE, results = "asis"}
number_of_bins_selected <- c(20,30)
i <- 1 # If for loop commented
#for (i in 1:length(number_of_bins_selected)) {
cat("## New slide \n")
input1_id <- paste(i, "nbreaks", sep = "_")
input2_id <- paste(i, "bw_adjust", sep = "_")
inputPanel(
selectInput(input1_id, label = "Number of bins:",
choices = c(10, 20, 35, 50), selected = number_of_bins_selected[i]),
sliderInput(input2_id, label = "Bandwidth adjustment:",
min = 0.2, max = 2, value = 1, step = 0.2)
)
renderPlot({
hist(faithful$eruptions, probability = TRUE, breaks = as.numeric(input[[input1_id]]),
xlab = "Duration (minutes)", main = "Geyser eruption duration")
dens <- density(faithful$eruptions, adjust = input[[input2_id]])
lines(dens, col = "blue")
})
cat(" \n\n") # End the slide
#}
```
Isn't it possible to create slides on a for loop ?

Xtable grey rows overwritting vertical lines

---
title: "Title"
author: ''
date: ''
output:
pdf_document:
template: default.tex
geometry: top=0.5cm, bottom=0.5cm, left=0.5cm, right=0.5cm
header-includes: null
fontsize: 4pt
classoption: portrait
sansfont: Calibri Light
---
#Name1: `r "Name1"`
#Name2: `r "Name2"`
```{r, echo=FALSE, message=FALSE, warning=FALSE, results='asis'}
df <- mtcars
n = nrow(df)
hlines=c(-1,0,(n-1),n)
my_align = "c|c|c|ccccc|ccc|c|"
rws <- seq(1, (n-1), by = 2)
col <- rep("\\rowcolor[gray]{.90} ", length(rws))
xtable::print.xtable(xtable(df
, align = my_align)
, add.to.row = list(pos = as.list(rws), command = col)
, booktabs = F
, hline.after = hlines, type = "latex")
```
I am using an Rmarkdown to print a table which has a lot of formatting. When I add the add.to.rwo part to get grey and white alternate rows the vertical lines are removed in the grey rows.
How do I correct this? It is very difficult to create a reproducible example but hopefully the same problem will apply to any df (with the correct Latex packages behind it)
Thanks :)
Try comparing these two tables. The first is your table as you coded it, the second is done by pixiedust with the hhline option set to TRUE.
---
title: "Title"
author: ''
date: ''
output:
pdf_document:
geometry: top=0.5cm, bottom=0.5cm, left=0.5cm, right=0.5cm
header-includes:
- \usepackage{amssymb}
- \usepackage{arydshln}
- \usepackage{caption}
- \usepackage{graphicx}
- \usepackage{hhline}
- \usepackage{longtable}
- \usepackage{multirow}
- \usepackage[dvipsnames,table]{xcolor}
fontsize: 4pt
classoption: portrait
sansfont: Calibri Light
---
#Name1: `r "Name1"`
#Name2: `r "Name2"`
```{r, echo=FALSE, message=FALSE, warning=FALSE, results='asis'}
library(xtable)
df <- mtcars
n = nrow(df)
hlines=c(-1,0,(n-1),n)
my_align = "c|c|c|ccccc|ccc|c|"
rws <- seq(1, (n-1), by = 2)
col <- rep("\\rowcolor[gray]{.90} ", length(rws))
xtable::print.xtable(xtable(df
, align = my_align)
, add.to.row = list(pos = as.list(rws), command = col)
, booktabs = F
, hline.after = hlines, type = "latex")
```
```{r}
library(pixiedust)
dust(df,
hhline = TRUE,
keep_rownames = TRUE) %>%
medley_bw() %>%
sprinkle_colnames(.rownames = "") %>%
sprinkle(cols = c(".rownames", "mpg", "cyl", "qsec", "gear", "carb"),
border = "right") %>%
sprinkle(rows = nrow(mtcars),
border = "top") %>%
sprinkle(bg_pattern_by = "rows")
```