To get started with showyourwork!, install the latest version using the pip command:

pip install -U showyourwork

You’ll also need to have the conda package manager installed, which you can download from this url.

Setup a new article#

To set up a blank article, simply execute the showyourwork setup command. This command accepts one positional argument: the “slug” of the GitHub repository for the article, which is a string in the form user/repo. For example, to create a repository called article under the GitHub user rodluger, run

showyourwork setup rodluger/article

This will bring up the following prompt:

Let's get you set up with a new repository. I'm going to create a folder called


in the current working directory. If you haven't done this yet, please visit


at this time and create an empty repository called


As requested, if you haven’t yet created the remote repository, go to github.com/new in your browser to create an empty repository of the same name. There’s no need to create a README, gitignore file, or LICENSE at this time, as showyourwork! will set those up for you.

Press any key to bring up the next prompt:

You didn't provide a caching service (via the --cache
command-line option), so I'm not going to set up remote caching for this repository.

The showyourwork! workflow automatically caches the results of intermediate steps in your pipeline on Zenodo, but only if the user requests it via the --cache option. Let’s not worry about this for now – you can read up on how this all works at Zenodo integration.

Press any key to bring up the final prompt:

You didn't provide an Overleaf project id (via the --overleaf command-line
option), so I'm not going to set up Overleaf integration for this repository.

The workflow can also manage integration with an Overleaf project if that’s how you prefer to write your TeX articles. To enable this, re-run showyourwork setup with --overleaf=XXX, where XXX is the 24-character project id of a blank Overleaf repository (you can grab this value from the URL of the project after creating it on Overleaf). Let’s again skip this for now – read all about it at Overleaf integration.

Finally, press any key to generate the repository. This will create a new folder in the current working directory with the same name as your repo (article, in the example above) and set up git tracking for it.

Build locally#

Your new repository is instantiated from a bare-bones template with the minimal showyourwork! layout, which you can read about at Repository layout. The first thing you might want to do is edit the main TeX file, which is located at src/tex/ms.tex. Give it a custom title, edit the author name, and replace some of the Lorem Ipsum placeholder text with something informative. Then, compile the article PDF by running

showyourwork build

or, as a shorthand, just


in the top level of your repository. The first time you run this, showyourwork! will set up a new conda environment and install a bunch of packages, so it could take one or two minutes. The workflow will then build your article PDF, which by default is saved as ms.pdf in the top level of your repository:


There’s not much to see – mostly placeholder text. One thing to note, though, is the blue GitHub icon in the right margin next to the abstract: this is a hyperlink pointing to your github repository (or, in this example, to github.com/rodluger/article.)

Let’s turn this into a proper scientific article by adding a figure. In showyourwork!, all figures should be programmatically generated, so we start by creating a script to generate the figure. For simplicity, let’s create a script random_numbers.py to generate and plot some random numbers:

import matplotlib.pyplot as plt
import numpy as np
import paths

# Generate some data
random_numbers = np.random.randn(100, 10)

# Plot and save
fig = plt.figure(figsize=(7, 6))
fig.savefig(paths.figures / "random_numbers.pdf", bbox_inches="tight", dpi=300)

By default, the showyourwork! workflow expects figure scripts to be located in (or nested under) src/scripts, so that’s where we’ll put this script.


The default location for figure output (i.e., the generated .pdf figure files) is in the src/tex/figures directory, so we need to make sure figure scripts save their output into that folder, regardless of where the script is executed from. The simplest way to do this is to import the paths module, a file that is automatically included in the src/scripts directory when you create a new article repository with showyourwork!. This module defines a few convenient paths, like figures and data. These are instances of pathlib.Path objects pointing to the absolute paths of various useful workflow directories.

Now that we’ve created our figure script, let’s include the figure in our article by adding the following snippet in the body of src/tex/ms.tex:

            Plot showing a bunch of random numbers.

Here we’re using the standard figure environment and \includegraphics command to include a PDF in our article. The one important bit of syntax that is specific to showyourwork! is the \script command, which is how we tell showyourwork! that the figure src/tex/figures/random_numbers.pdf can be generated by running the script src/scripts/random_numbers.py. Note that within the \script command, all paths are relative to src/scripts (where the workflow expects these scripts to be located); within calls to \includegraphics and other similar commands, paths are relative to the graphicspath, which by default is src/tex/figures.


Previous versions of showyourwork! inferred the name of the figure script directly from the figure \label command. This functionality is now deprecated; users must now either use the \script command or define a custom Snakemake rule to generate a figure from a script.

If we now run showyourwork! again, we’ll get a message saying conda needs to download and install some more packages. Once that’s done, a message will inform us the figure random_numbers.pdf is being built, and if that goes well, we’ll get a recompiled article PDF that looks like this:


In addition to automatically building our figure for us, showyourwork! has also included a GitHub icon in the margin next to its caption, which points to the script that generated it (in this case, random_numbers.py). Importantly, the link points to the exact version of the script (i.e., to the specific commit SHA on GitHub) that was used to generate the figure.

If you haven’t yet pushed your changes to GitHub, that URL won’t exist yet; so let’s sync our changes with the remote next.

Build on the remote#

Whenever you make a change to your article (add text, add a figure, edit a script), make sure to git add any new/modified files, commit your changes, and then push to the GitHub remote:

git add src/scripts/random_numbers.py
git add src/tex/ms.tex
git commit -m "added a new figure"
git push -u origin main


Note that we’re only adding the figure script, not the figure file, to the list of files tracked by git. In fact, if you try to add the figure file, you’ll get an error:

git add src/tex/figures/random_numbers.pdf

The following paths are ignored by one of your .gitignore files:
Use -f if you really want to add them.

As soon as you push your changes to GitHub, a GitHub Action will be triggered on your repository, which will build your article from scratch on the cloud. To track the build, click on the actions tab of your repository. The first time your article is built, the action will have to download and install conda, so it will likely take a few minutes. Subsequent builds take advantage of intelligent cross-build caching, so they will likely run faster.

When the build is done, you can click on any of the badges on the front page of your repository:


these will take you to the build logs, the article tarball (containing the TeX files and all generated figures), and the compiled PDF of the article, respectively.

Build someone else’s article#

If you would like to build an article created by someone else with showyourwork!, simply clone the article’s GitHub repository:

git clone https://github.com/user/repo
cd repo

and, assuming you’ve installed showyourwork!, run the following command in the top level of the repository:

Setting up the workflow...
Generating the article PDF...

This will create a new conda environment for the workflow, install all the required dependencies, and run all of the scripts needed to generate the figures and results in the paper. At the end of the build, the article PDF will appear in the top level of the repository (usually called ms.pdf).

That’s it for this quickstart tutorial. Please check out the rest of the documentation for more information on how to customize your workflow, debug issues, etc.