This project is no longer being actively maintained. Feel free to use the code as-is, but please note that future updates or support are not planned.
Kumquat is a reporting tool for rendering RMarkdown files inside Rails.
RMarkdown, or .Rmd files are Markdown files with code chunks of R in them.
You can use Kumquat to integerate R output into Rails, which is useful for using libaries like ggplot2 in email reporting tools.
For example, consider a typical render call to a partial:
render '_a_knitr_report'
This partial is _a_knitr_report.Rmd, a regular RMarkdown file stored in your app:
A Test Report for Kumquat
========================================================
This is some code for Knitr, including [a link](https://www.kickstarter.com).
A list:
* One
* Another thing
* Last thing!
### More stuff.
Lorem Ipsum.
```{r, fig.width=10, fig.height=8, echo=FALSE, message=FALSE}
library(ggplot2)
qplot(data = data.frame( x = runif(100), y = runif(100) ), x = x, y = y)
```Which yields the following output:
Kumquat can be integrated anywhere inside Rails views where partials are called.
Kumquat requires R to be installed on the machines running your application. The main action is then in Knit2HTML which shells out to Rscript and runs the necessary knitr package methods to compile HTML.
When you install Kumquat a template handler is registered to handle Rmd files.
A mail interceptor, KumquatReportInterceptor is also registered to attach images inline to Rmd reports marked with an X-KUMQUAT signature which you can add to a mail object accordingly:
mail :to => @admin.email,
:reply_to => "kumquat@example.com",
:subject => "Test Kumquat Report",
'X-KUMQUAT' => true do |format|
format.text { render layout: nil }
format.html { render layout: nil }
endKumquat requires R to be installed, and the knitr packages to render RMarkdown files. The rest of the R dependencies are up to you, but need to be specified in the KNITR_LIBRARIES constant in the Knit2HTML class.
If you're curious about installing R using Chef, check out a suggested recipe on the wiki
If you need to connect R to a database and provide credentials from your application, you can do so in an application initializer like so:
Kumquat.database_config(Configs[:redshift].merge({ database_connector: "RPostgres::Postgres()" }))We are using Hadley Wickham's RPostgres library which can be installed in R accordingly:
install.packages("devtools")
devtools::install_github("RcppCore/Rcpp")
devtools::install_github("rstats-db/DBI")
devtools::install_github("rstats-db/RPostgres")
library(DBI)If you'd like to deliver emails in your local environment, make sure your config.action_mailer.delivery_method is configured properly, and you have a way to send email locally (on OS X, run the command sudo postfix start).
You can view kumquat and email logs for debugging tail -f log/kumquat.log
