Passing parameters through an RMarkdown document? - r

Question

I am looking to create unique/individual reports for a given list of vendors. The ideal output format would be a seperate html file of the given vendors information.
The issue is that I am having trouble wrapping my head around creating parameterized reports in RMarkdown. I have been taking a look at this link to understand how to loop/iterate through RMarkdown reports
To illustrate and share the logic of what I want to execute is the following:
for (vendor in vendor.name) {
rmarkdown::render('input.Rmd', params = list(vendor = vendor))
}
Where I then have a print out of:
Vendor-1.html, Vendor-2.html, …, Vendor-4.html, and vendor-4.html
Which then gets saved locally to my computer. The part I have wrapping my head around is say we have a bar graph of sales by month, how would the parameter passing through the entire document know when to change the the vendor number for a unique view.
If anyone can share an example either iris, mtcars, or any base dataset within R I would really appreciate it. Looking at how this workflow/logic would work because I am struggling to understand the concept.
To specify, say I have this chunck of code here. How would the params$vendor function know to loop over another vendor if I am not calling it within the chunk. Within my dplyr verbs on my filter should I do , MVNDR_NBR == qc_sales$vendor_number or params$vendor? This is the piece that has me most confounded
sales_2021_stock <- qc_sales %>%
filter(FSCL_YR == 2021
, STR_NBR != '8119'
, MAPPED_ORD_SRC == 'QC'
, so_flg == 0
, YTLW_TY_LY_FLG == 'TY'
, MVNDR_NBR == '60031167'
, !SUB_DEPT_NBR %in% c('0025','0028')) %>%
group_by(MVNDR_NBR, MVNDR_NM, FSCL_WK_NBR, FSCL_YR) %>%
summarise(Sales = sum(ESVS_NET_SLS)) %>%
mutate(FSCL_YR = as.character(FSCL_YR)) %>%
collect()
sales_2020_stock <- qc_sales %>%
filter(FSCL_YR == 2020
, STR_NBR != '8119'
, MAPPED_ORD_SRC == 'QC'
, YTLW_TY_LY_FLG == 'LY'
, so_flg == 0
, MVNDR_NBR == '60031167'
, !SUB_DEPT_NBR %in% c('0025','0028')) %>%
group_by(MVNDR_NBR, MVNDR_NM, FSCL_WK_NBR, FSCL_YR) %>%
summarise(Sales = sum(ESVS_NET_SLS)) %>%
mutate(FSCL_YR = as.character(FSCL_YR)) %>%
collect()
sales_comp_line_stock <- rbind(sales_2021_stock, sales_2020_stock)
stock_comp <- ggplot(sales_comp_line_stock, aes(x = FSCL_WK_NBR, y = Sales, color = FSCL_YR ))+
geom_line(size = 1.25, aes(color = FSCL_YR))+
geom_smooth(size = .50, aes(color = FSCL_YR), se = FALSE, method = "auto")+
scale_x_continuous(breaks=seq(0,weeks,1))+
scale_y_continuous(labels = scales::dollar_format(scale = .0001, suffix = "K"))+
scale_color_manual(values=c("#0298F9", "#F96302"))+
theme_economist()+
ggtitle('Week-Over-Week Sales (Stock) 2021 v 2020')+
theme(
panel.grid.major = element_line(linetype = "dotted"),
axis.text = element_text( size = 10),
legend.position = c(0, 1),legend.justification = c(0, 1),
plot.title = element_text( size = 14, margin=margin(0,0,20,0), hjust = 0.5),
panel.background = element_rect(fill = NA),
strip.text = element_text(size=10)
)
stock_comp
My dataset looks as follows, these are the same vendors included in my parameter, how would I create a ggplot displaying sales month over month printed into an individual html output?:
Ideally one plot would be written like this:
ggplot(sample_vendor_tbl, aes(x = FSCL_MTH_NM, y = Sales)) +
geom_col()
To print out multiple ggplots within a document I would do the following:
for (i in vendor_nbr){
ggplot(mydata, aes(x = Month, y = sales))+
geom_col()
}
I am just confused here when we need to accounting for the parameter. How can I create a plot for the given vendor for its print out, similar to the example posted in your answer. I basically want to do exactly what you did in your answer with the charts but leveraging ggplot instead of base R
To create a ggplot using parameters I had to pull in params$mvndr_nbr into my dplyr verbs as follows:
sample_vendor_tbl %>%
filter(MVNDR_NBR == params$MVNDR_NBR) %>%
ggplot(aes(x = FSCL_MTH_NM, y = Sales)) +
geom_col()

Answer 1

I think the step you were missing was specifying the output filename so that each "vendor" would have its own file; otherwise, the same filename is overwritten each time leaving you with a single HTML document.
An example:
---
title: mtcars cyl
author: r2evans
params:
cyl: null
---
We have chosen to plot a histogram of `mtcars` where `$cyl` is equal to `r params$cyl`:
```{r}
dat <- subset(mtcars, cyl == params$cyl)
if (nrow(dat) > 0) {
hist(dat$disp)
}
```
Calling this with:
for (cy in c(4,6,8)) {
rmarkdown::render("~/StackOverflow/10466439/67525642.Rmd",
output_file = sprintf("cyl_%s.html", cy),
params = list(cyl = cy))
}
will render three HTML files, cyl_4.html, cyl_6.html, and cyl_8.html, each with differing content: