在 RMarkdown 中使用用户输入重复块

Question

I am attempting to make an automated script to conduct analysis of Sanger Sequencing data. My end goal is to convert my R script into a knittable R Markdown file displaying chromatograms and what not. My current hang up in the transition to an R Markdown file is in using user input in repeating loops. I'm aware that the readline command doesn't work while knitting as this is not an interactive process. From what I understand, the only way to receive user input while knitting is through the use of parameters. However, the nature of my current script is such that the user specifies how many files are being analyzed which then decides how many times certain loops are executed. As one might imagine, lacking user input results in infinite loops. This is an example from my current script that runs just fine in RStudio:

# install.packages("BiocManager")
# BiocManager::install("sangerseqR")
# BiocManager::install("Biostrings")

library(BiocManager)
library(sangerseqR)
library(Biostrings)

refnum <- readline(prompt = "Enter Number of Reference Files to be aligned: ")
refnum <- as.integer(refnum)
c <- 1
repeat{
  # Compiles reference files
  repeat{
    wtreffile <- readline(prompt = paste("Enter WT Reference File ", c, ": ", sep = ""))
    if (!file.exists(wtreffile)) {
      message("File Does Not Exist. Please Re-Enter.")
    } else {
      break
    }}
  repeat{
    insreffile <- readline(prompt = paste("Enter INS/DEL/MUT Reference File ", c, ": ", sep = ""))
    if (!file.exists(insreffile)) {
      message("File Does Not Exist. Please Re-Enter.")
    } else {
      break
    }}
  # Reads sequence from reference files and writes it as a DNAString object to be used for alignment
  wt <- readDNAStringSet(wtreffile, format = "fasta", nrec = 1L, skip = 0L, seek.first.rec = TRUE)
  ins <- readDNAStringSet(insreffile, format = "fasta", nrec = 1L, skip = 0L, seek.first.rec = TRUE)
  # Aligns wild type and insert reference files pairwise
  pa <-pairwiseAlignment(pattern = ins, subject = wt)
  pafile <- readline(prompt = paste("Enter name for reference alignment of ", wtreffile, " and ", insreffile, ": ", sep = ""))
  writePairwiseAlignments(pa, file = pafile)
  c <- c + 1
  if (c > refnum / 2) {
    break
  }
}

As you can probably see, the amount of times this chunk repeats is entirely dependent on user input. My big question is whether its possible to accomplish this same task (allowing the user to specify the amount of files to be used and select which files to use based on that amount) in RMarkdown using parameters, or if I'll have to sacrifice some automation in favor of user customization. Is it possible to create parameters in RMarkdown based on the input of other parameters? Let me know if anything needs cleared up and thanks in advance!!!

Answer 1

Expanding on my comment, if you don't need interactivity, you can also pass the chosen parameters to rmarkdown::render in a regular R script, obviously.

Here is a simple example, passing the parameters x and fn to the report:

This is the report file report.rmd saved in the working directory:

title: "Report"
output: pdf_document
params:
  x: NA
  fn: NA
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## Plot with parameters passed from R script

The red line shows the `r params$fn` of the distribution.

```{r plot, echo=FALSE}
x <- params$x
fn <- params$fn

set.seed(1)
df <- rnorm(100)
plot(density(df))
if(x!="NA") abline(v=x, col="red") 

# in this example, it would make more sense to pass the function, not `x`:
# if(x!="NA") abline(v=do.call(fn, list(df)), col="red") 
```

This is where the parameters are passed to the report in an R script:

tempReport <- file.path(tempdir(), "report.Rmd")
file.copy("report.Rmd", tempReport, overwrite = TRUE)

# parameters passed to report:
params <- list(x = 0.3, fn = "mean") 
rmarkdown::render(tempReport, output_file = "~/Downloads/new.report.pdf", run_pandoc=TRUE,
                  params = params, output_format="pdf_document", clean=FALSE,
                  envir = new.env(parent = globalenv())
)