gofigR

gofigR is the R client for https://gofigr.io, a zero-effort reproducibility engine. It works with any R library which outputs to R graphics devices, but automatic figure detection and publishing is currently limited to ggplot.

Compatibility

gofigR integrates with R markdown, both in knitr and in interactive sessions in RStudio. GoFigr also works in scripts. We tested with R 4.3.2 but any reasonably recent version should work.

GoFigr will automatically publish all ggplot output assuming you call gofigR::enable(auto_publish=TRUE). GoFigr will not publish old-style R plots unless you tell it to. See the “Usage” section below.

Installation

install.packages("gofigR")

You can also install the development version from GitHub:

library(devtools)
devtools::install_github("https://github.com/gofigr/gofigR")

Configuration

On the R prompt, simply load the gofigR package and call gfconfig(). You only need to do this once.

If you don’t have an account, you can register at https://app.gofigr.io/register.

> library(gofigR)
Attaching package: ‘gofigR’

> gfconfig()
-------------------------------------------------------------------
Welcome to GoFigr! This wizard will help you get up and running.
-------------------------------------------------------------------

Username: demo
Password: *******************
Testing connection...
  => Success
API key (leave blank to generate a new one): 
Key name (e.g. Alyssa's laptop): work laptop
Fetching workspaces...


| Number|Name              |Description                  |API.ID                               |
|------:|:-----------------|:----------------------------|:------------------------------------|
|      1|Primary Workspace |demouser's primary workspace |98f328fc-f984-482c-b7f6-8ed272527a42 |
|      2|Rocketship Bio    |Let's do some science        |bff6c952-fb2c-4333-80e5-5dc7291d47cc |
|      3|Plotly demos      |N/A                          |1424ea0f-7d42-4ef5-9ba8-41aa9c5b1a94 |

Please select a default workspace (1-3): 1

Configuration saved to /Users/maciej/.gofigr. Happy analysis!

> 

Usage

To enable GoFigr, simply call enable in your setup chunk. analysis_name specifies the analysis under which all figures will be published (it will be created automatically if it doesn’t exist).

```{r setup, include=FALSE}
library(gofigR)

gofigR::enable(analysis_name="My first Rmd analysis",
               auto_publish=TRUE)
```

auto_publish is FALSE by default. Set it to TRUE to override plot and print and publish figures automatically.

After calling enable you can knit your markdown as-is. However, you can also customize GoFigr through chunk options:

• gofigr_figure_name: manually specify the name of the figure
• gofigr_on: set to FALSE to disable GoFigr within a chunk

For example:

```{r pressure, echo=FALSE, gofigr_figure_name="My first figure"}
plot(pressure)
```
  
```{r pressure2, echo=FALSE, gofigr_on=FALSE}
# Won't be published
plot(pressure)
```

Automatic output capture

If auto_publish is on, GoFigr will intercept all calls to plot and print and publish the results if they are from a compatible library (at the moment, only ggplot).

Manual capture

To capture output from old-style R plotting or from other libraries, including when the plot is built iteratively across multiple expressions, wrap your code in gofigR::capture:

gofigR::capture({
  plot(pressure, main="Pressure vs temperature")
  text(50, 50, "My pretty figure")
}, data=pressure)

This is handy when you build a plot iteratively. For example, you may call plot(...) first, followed by a call to text() to add annotations, or legend() to place the legend.

Note the optional data argument following the expression. It specifies the data which you want to associate with the figure – it will show up under “files” (as .RDS) once published.

Adding support for other plotting libraries

If you have a plotting function which you use often and which you would like to auto-publish, you can use intercept:

barplot <- gofigR::intercept(graphics::barplot)

Subsequent calls to barplot will then automatically publish to GoFigr.

Interactive use

gofigR works best with knitr, but interactive sessions within RStudio are also supported.

When running within RStudio, you will see both the original plots as well as their published & watermarked counterparts.