Generate beautiful, customizable and reproducible Markdown reports. Please note that this app is "beta", but fully functional already.
Try a simplified live demo here: https://www.nilsole.net/rmarkdownreportmanager_app/index.html
The live demo is highly restricted. Only report generation and asset downloading are permitted for demo users.
- Include R tibbles as input datasets (RDS file format expected).
- Use customizable R packages ("report packages") as report templates (Rmarkdown format expected).
- Generate reports: One report file per input dataset row (mapping dataset rows to output files). HTML, PDF and Word supported.
The app turns datasets (input) into report documents (output). It does this by converting each line of the input dataset into a document (or several documents). To make the process reproducible, the app keeps track of datasets and templates carefully, helping the user to manage even large numbers of reports.
As of now, input datasets are expected be RDS export files of tibbles.
How the report will look like is defined by a "report package", a conventional R package that will render the reports and is usually expected to contain rmarkdown templates. RStudio provides a simple guide to creating custom R packages.
The report manager makes it all work as follows: It slices the input dataset into rows and then calls the function makeReport, which is expected to be included in the report package.
Here is one example of such a function. data is a tibble that contains one row, whereas folder_path is the path of the output folder.
makeReport <- function( data, folder_path ) {
input_file <- file.path(".","inst","rmarkdown","templates","my_template","skeleton","skeleton.Rmd")
output_filename <- paste0(formatC(data$meta_rownr, width = 10, format = "d", flag = "0"),".html")
tf <- tempfile()
dir.create(tf)
rmarkdown::render(input = input_file,
output_file = output_file,
intermediates_dir = tf,
output_dir = folder_path,
params = as.list(data),
quiet = TRUE)
unlink(tf)
}
The function hands the data over to the rmarkdown template as parameters. In addition to the information included in the input dataset, the report manager adds checksum information about the procedure to the dataset (with a "meta_" prefex). By doing so, it makes it easier to identify unique report documents and their properties.
---
title: "`r params$title`"
params:
date_time: !r Sys.time()
meta_config_md5sum: NA
meta_dataset_md5sum: NA
meta_dataset_name: NA
meta_reportpackage_md5sum: NA
meta_reportpackage_name: NA
meta_reportpackage_version: NA
meta_rownr: NA
title: "Some usual title"
---
## Some headline
The document contents goes here. Parameterized information can be included using markdown placeholders. The document may also contain charts, plots and even custom R code.
---
<p style="font-size:8pt;">Meta Information</p>
meta_dataset_name: `r params$meta_dataset_name`<br/>
meta_dataset_checksum: `r params$meta_dataset_md5sum`<br/>
meta_reportpackage_name: `r params$meta_reportpackage_name`<br/>
meta_reportpackage_version: `r params$meta_reportpackage_version`<br/>
meta_reportpackage_checksum: `r params$meta_reportpackage_md5sum`<br/>
meta_config_checksum: `r params$meta_config_md5sum`<br/>
meta_rownr: `r params$meta_rownr`<br/>
Document rendered: `r params$date_time`
| Description | Status | Issue |
|---|---|---|
| Let report generator work in parallel | Planned | related |
| Make configuration for report generation more customizable (file formats, ...) | Planned | |
| Include preview for datasets and reports | Planned | |
| Improve reproducibility with regard to R package versions (possibly by adding Packrat support) | Planned | |
| Add user management (rights & roles) | Planned | |
| Let users import datasets from other sources (databases, csv files, ...) | Planned | |
| Improve validation feedback when creating new report package | Planned | |
| Improve documentation (including "best practices") | Planned |
- Prepare your R environment:
- Clone git repo (e.g. import as an R package in RStudio).
- Create R source package.
- Deploy R package on Shiny server (please note that this app will currently not work on shinyapps.io, since it requires local data persistence).
- Update your shiny-server.conf accordingly.
- Make sure that the server user ("shiny" by default) of your Shiny server has sufficient rights to create, delete and modify files in the app directory. Please take a look at the Shiny server manual for further information.
It is also recommended to run the package tests to make sure everything is installed correctly:
library(devtools)
library(rmarkdownreportmanager)
devtools::test()
After installation, the app will allow you to apply restrictions to the user rights. Open the tab "Configuration" to change these settings.
You will be expected to enter valid JSON code. As soon as you click on the button "Update config file" and your input is accepted, the configuration will be updated and the app will shut down. Restart the app to make the changes take effect.
By setting the property update_config to false, you can prevent user from making further changes. The configuration tab will then disappear after you restart the app. Please note that you will not be able to undo this step unless you change the contents of the file ./config/config.json.