Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

add post from njtierney.com on magic reprex; closes #101 #103

Merged
merged 1 commit into from
Sep 18, 2017
Merged

add post from njtierney.com on magic reprex; closes #101 #103

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
171 changes: 171 additions & 0 deletions vignettes/articles/magic-reprex.Rmd
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,171 @@
---
title: "Magic reprex"
author: "[Nicholas Tierney](njtierney.com)"
---

> This article was originally posted on [Nick Tierney's blog](https://www.njtierney.com/post/2017/01/11/magic-reprex/), and has been contributed here to help spread the good word about `reprex`! :) - **Nick Tierney**

Making reproducible examples can be hard.

There's a lot of things you need to consider. Like, making sure your environment is clean, the right packages are loaded, the code is formatted nicely, and images are the right resolution and dimension.

Getting all of these ducks lined up can sometimes take a couple of minutes, if you have a nice tightly defined problem. Other times, it can take much, much longer. Say for example, if you had a variable loaded in your environment that you forgot how to create, and you didn't realise it was so important and OH GOD HOW DID I EVEN DO THIS? Ahem.

So, after some sensible swearing and frustration, you might ask yourself something like:

> Surely there's an easy, reproducible way to ... make reproducible examples?

Enter the R package `reprex` [[CRAN](https://cran.r-project.org/web/packages/reprex/index.html)] [[GitHub](https://github.com/tidyverse/reprex)], by [Jenny Bryan](http://www.github.com/jennybc).

In the fashion of the times right now, I'll illustrate `reprex` with a gif, which shows the following things:

1. Take your code, copy it to the clipboard.
2. Run `reprex`, wait a moment for the code to execute.
3. `reprex` creates a html preview of your code.
4. A markdown chunk is also copied to the clipboard.

<p>
<div style="height: auto">
<img src="https://njtierney.updog.co/gifs/reprex-copy.gif" alt="Gif of reprex copying code, running reprex, and demonstrating the html preview of the code text" style="width: 100%;max-height: 100%" />
</div>
</p>


This means you can copy this to say, GitHub (`reprex(venue = "gh")`) (the default) to help you file an issue:


<p>
<div style="height: auto">
<img src="https://njtierney.updog.co/gifs/reprex-github.gif" alt="Gif of reprex copying code, running reprex, and demonstrating the html preview of the code text" style="width: 100%;max-height: 100%" />
</div>
</p>

There is also an option for Stack Overflow(`reprex(venue = "so")`.

How does this work? `reprex` takes the section of copied code and then runs `rmarkdown::render` on it, which starts a fresh environment, and runs the code.
This means that `reprex` will fail hard and fail early if your examples aren't reproducible. If you don't have the correct library loaded, it will show the error messages.

<p>
<div style="height: auto">
<img src="https://njtierney.updog.co/gifs/reprex-fail-1.gif" alt="Gif of reprex copying code, running reprex, and demonstrating the html preview of the code text" style="width: 100%;max-height: 100%" />
</div>
</p>

But then when you fix that, by adding library calls to [`visdat`](https://github.com/njtierney/visdat), [`ggplot2`](https://github.com/tidyverse/ggplot2) and [`naniar`](https://github.com/njtierney/naniar), and you get what you want:

<p>
<div style="height: auto">
<img src="https://njtierney.updog.co/gifs/reprex-success-1.gif" alt="Gif of reprex copying code, running reprex, and demonstrating the html preview of the code text using visdat, naniar, and ggplot2 packages" style="width: 100%;max-height: 100%" />
</div>
</p>

So you know that when you run `reprex`, and it gives you a reproducible example, it will work!

Also worth mentioning is that rather magically, `reprex` also handles images. I didn't even notice this at first. It was just one of those things where "it just works". `reprex` handles this by uploading them to imgur using the `knitr::imgur_upload` function, which creates markdown links to the images.

so

```r

library(tidyverse)

# run a lm

lm_fit <- lm(Sepal.Length ~ . , data = iris)

# use broom to clean it up

library(broom)

lm_augmented <- augment(lm_fit)

head(lm_augmented)

# Explore the fit of the model

ggplot(data = lm_augmented,
aes(x = .fitted,
y = Sepal.Length)) +
geom_point()

```

gets turned into


```
library(tidyverse)
#> Loading tidyverse: ggplot2
#> Loading tidyverse: tibble
#> Loading tidyverse: tidyr
#> Loading tidyverse: readr
#> Loading tidyverse: purrr
#> Loading tidyverse: dplyr
#> Conflicts with tidy packages ----------------------------------------------
#> filter(): dplyr, stats
#> lag(): dplyr, stats

# run a lm

lm_fit <- lm(Sepal.Length ~ . , data = iris)

# use broom to clean it up

library(broom)

lm_augmented <- augment(lm_fit)

head(lm_augmented)
#> Sepal.Length Sepal.Width Petal.Length Petal.Width Species .fitted
#> 1 5.1 3.5 1.4 0.2 setosa 5.004788
#> 2 4.9 3.0 1.4 0.2 setosa 4.756844
#> 3 4.7 3.2 1.3 0.2 setosa 4.773097
#> 4 4.6 3.1 1.5 0.2 setosa 4.889357
#> 5 5.0 3.6 1.4 0.2 setosa 5.054377
#> 6 5.4 3.9 1.7 0.4 setosa 5.388886
#> .se.fit .resid .hat .sigma .cooksd .std.resid
#> 1 0.04479188 0.09521198 0.02131150 0.3077918 3.570856e-04 0.31367294
#> 2 0.05514933 0.14315645 0.03230694 0.3076565 1.251718e-03 0.47429637
#> 3 0.04690495 -0.07309695 0.02336968 0.3078349 2.317688e-04 -0.24106926
#> 4 0.05135928 -0.28935683 0.02801904 0.3069173 4.396118e-03 -0.95656076
#> 5 0.04736842 -0.05437691 0.02383379 0.3078627 1.309299e-04 -0.17937436
#> 6 0.05592364 0.01111427 0.03322050 0.3078956 7.772793e-06 0.03684044

# Explore the fit of the model, looking

ggplot(data = lm_augmented,
aes(x = .fitted,
y = Sepal.Length)) +
geom_point()

`![](http://i.imgur.com/h72ZEKG.png)`
```

You can even show the results of the `session_info`, for maximum reproducibility, by adding the argument: `reprex::reprex(si = TRUE)`, which adds a super nifty drop down box for the session info.


<p>
<div style="height: auto">
<img src="https://njtierney.updog.co/gifs/reprex-si.gif" alt="Gif of reprex copying code, running reprex, showing option SI = TRUE" style="width: 100%;max-height: 100%" />
</div>
</p>

I really like these sorts of small tools in R. Squinting at a distance you might make the mistake of thinking:

> so what, it does this __one little thing__

But then when you're actually up close, you realise this means that you don't have to bend your brain around doing this thing, which can range from mildly annoying to mostly infuriating.

You save time and energy, and can focus more on **doing what you want to do**, and less on **fighting against interruptions**. This comic by [Jason Heeris](http://heeris.id.au/2013/this-is-why-you-shouldnt-interrupt-a-programmer/) illustrates this point nicely, I think:

<p>
<div style="height: auto">
<img src="https://njtierney.updog.co/img/ProgrammerInterrupted.png" alt="Why you shouldn't interrupt a programmer" style="max-height: 100%"/>
</div>
</p>


Once again thank you to Jenny Bryan for making our lives easier. Jenny will be joining the RStudio team this year and I am really looking forward to seeing what she gets up to.

Read more about `reprex` on github [here](http://www.github.com/tidyverse/reprex).