https://github.com/coolbutuseless/ggreverse

Reverse a ggplot object back into code
https://github.com/coolbutuseless/ggreverse

Last synced: 3 months ago
JSON representation

Reverse a ggplot object back into code

• Host: GitHub
• URL: https://github.com/coolbutuseless/ggreverse
• Owner: coolbutuseless
• License: mit
• Created: 2019-04-26T07:45:35.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2019-05-31T21:27:31.000Z (over 5 years ago)
• Last Synced: 2024-10-12T21:24:57.596Z (3 months ago)
• Language: R
• Size: 55.7 KB
• Stars: 68
• Watchers: 4
• Forks: 5
• Open Issues: 3
• Metadata Files:
  • Readme: README.Rmd
  • License: LICENSE

Awesome Lists containing this project

• awesome-r-dataviz - ggreverse - Reverse a ggplot object back into code. (ggplot / Miscellaneous)

README

        
---

output: github_document

---

```{r, include = FALSE}

suppressPackageStartupMessages({

  library(ggplot2)

})

knitr::opts_chunk$set(

  collapse = TRUE,

  comment = "#>",

  fig.path = "man/figures/README-",

  out.width = "100%",

  fig.height = 2

)

```

# ggreverse

![](https://img.shields.io/badge/Status-alpha-orange.svg)

![](https://img.shields.io/badge/Version-0.1.1-blue.svg)

`ggreverse` takes a ggplot object and returns the code to create that plot.

This package is written as a learning exercise to help me figure out 

the internal structure of a ggplot object.

## Releases

* `0.1.0` - initial release

* `0.1.1` - improved theme handling

## Installation

You can install from [GitHub](https://github.com/coolbutuseless/) with:

``` r

# install.packages("remotes")

remotes::install_github("coolbutuseless/ggreverse")

```

## Example `ggreverse::convert_to_code()`

1. Create a ggplot

2. Convert the ggplot back into code using `ggreverse`

3. Convert the code back into a plot

```{r}

library(ggreverse)

plot_df <- mtcars

# Create a ggplot2 plot object

p <- ggplot(plot_df) +

  geom_point(aes(mpg, wt, colour = cyl), size = 3) +

  labs(title = "hello") +

  theme_bw() + 

  theme(legend.position = 'none') + 

  coord_equal()

```

```{r echo = FALSE}

p

```

```{r echo=TRUE, eval=FALSE}

# Convert the plot object back into code

plot_code <- ggreverse::convert_to_code(p)

print(plot_code)

```

```{r echo=FALSE, eval=TRUE}

plot_code <- convert_to_code(p)

styler::style_text(

  gsub("[+]", "+\n", plot_code)

)

```

```{r}

# Parse the plot code back into a plot - which should match the original plot

eval(parse(text = plot_code))

```

## Technical Notes

* the `data` arguments to `ggplot()` and `geom()` are evaluated at call time. There is

  no easy way to recover the name of the data argument.

    * `ggreverse` tries to match the actual data in the ggplot object against a named

      object in the plotting environment. Otherwise it uses a generic data name

* aesthethic mappings are evaluated at call time, so tidyeval and `aes_string()` mappings

  are supported, but `ggreverse` will only include the final variable name mapping.

* Layers are currently extracted as `geom_x(stat = 'y')` rather than `stat_y(geom='x')`.  

  I'm not sure if there are any cases where these aren't equivalent.

## ToDo

* Extracting `facet` and `scales` information.

* Complete themes which are customisations of built-in themes could be 

  more compact if nested diffs where done between themes, rather than 

  just a `shallow_diff()`

* Lots of other stuff :)

## SessionInfo

Developed against:

* R 3.5.3

* ggplot2 v3.1.1