fslab.fsx

(**
---
category: Documentation
categoryindex: 1
index: 6
---
*)

(*** condition: prepare ***)
#nowarn "211"
#r "../src/RProvider/bin/Release/net7.0/RDotNet.dll"
#r "../src/RProvider/bin/Release/net7.0/RProvider.Runtime.dll"
#r "../src/RProvider/bin/Release/net7.0/RProvider.DesignTime.dll"
#r "../src/RProvider/bin/Release/net7.0/RProvider.dll"
(*** condition: fsx ***)
#if FSX
#r "nuget: RProvider,{{package-version}}"
#endif // FSX
(*** condition: ipynb ***)
#if IPYNB
#r "nuget: RProvider,{{package-version}}"
#endif // IPYNB

(**
Working with the FsLab ecosystem
===============================

The R type provider is interoperable with other packages in
FsLab through its plugin architecture. Some examples are shown
below. If you would like to see better interoperability between
R and other FsLab packages, submit an issue to their repository
for the creation of an RProvider plugin.

### Deedle - data frame manipulation

Deedle provides types for F# data frame and time series manipulation.
To use with RProvider, first install the Deedle.RPlugin package from
nuget; once this is installed, you do not need to reference it in your
script files.
*)

#r "nuget:Deedle.RPlugin"

(**
In a new F# script file, first open Deedle and RProvider: 

    [lang=fsharp]
    #r "nuget:RProvider"
*)

#r "nuget:Deedle"

open RProvider
open RProvider.``base``
open RProvider.datasets
open Deedle

(**
The Deedle R plugin should be loaded by the R type provider automatically.
You can now convert back and forth between R data frames and Deedle frames
by using type annotations:
*)

let mtcars : Frame<string, string> = R.mtcars.GetValue()

// Pass Deedle data to R and print the R output
R.as_data_frame(mtcars)

// Pass Deedle data to R and get column means
R.colMeans(mtcars)