cookiecutter-data-science/docs/docs/index.md

Cookiecutter Data Science

A logical, reasonably standardized, but flexible project structure for doing and sharing data science work.

Why use this project structure?

We often think of data analysis as just the resulting report, insights, or visualizations. Even though these end products are generated by code, it's easy to focus on making the products look nice and ignore the quality of the code that generates them. While these end products are generally the main event, code quality is still important! And we're not talking about bikeshedding the aesthetics or pedantic formatting standards — it's ultimately about correctness and reproducibility.

It's no secret that good analyses are often the result of very scattershot and serendipitous explorations, tentative experiments, and rapidly testing what works and what doesn't. That is all part of the process for getting to the good stuff, and there is no magic bullet to turn data exploration into a simple, linear progression. That being said, once started it is not a process that lends itself to thinking carefully about the structure of your code or project layout.

We think it's a pretty big win all around to let someone else do that up-front thinking and setup for you. Here's why:

Other people will thank you

Nobody sits around before creating a new Rails project to figure out where they want to put their views; they just run rails new to get a standard project skeleton like everybody else.

A well-defined, standard project structure means that a newcomer can begin to understand an analysis without digging in to extensive documentation. Well organized code is self-documenting and provides a lot of context for your code without much overhead. People will thank you for this because they can:

• Collaborate easily with you on this analysis
• Easily learn from your analysis about the process and the domain
• Feel confident in the conclusions the analysis presents

A consistent project structure means less random searching for what gets called where. A good example of this can be found in web development frameworks like Ruby on Rails, Django, and most others. Nobody sits around before creating a new Rails project to figure out where they want to put their views; they just run rails new to get a standard project skeleton like everybody else. And because the default project structure is reasonably logical and standard across most projects, it takes almost no time at all for somebody who has never seen a particular project to figure out where they would find the various moving parts.

Ideally, that's how it should be when a colleague opens up your data science project.

You will thank you

Ever tried to reproduce an analysis that you did a few months ago or even a few years ago? You may have written the code, but it's now impossible to decipher whether you should use make_figures.py.old, make_figures_working.py or new_make_figures01.py to get things done. Here are some questions we've learned to ask with a sense of existential dread:

• Are we supposed to go in and join the "region" column to the data before we get started or did that come from one of the notebooks?
• Come to think of it, which notebook do we have to run first before running the plotting code: was it "process data" or "clean data"?
• Where did the shapefiles get downloaded from for the geographic plots you made?
• Et cetera, times infinity.

These types of questions are painful and are symptoms of a disorganized project. A good project structure encourages practices that make it easier to come back to old work, for example separation of concerns, abstracting analysis as a DAG, and engineering best practices like version control.

Getting started

With this in mind, we've created a data science cookiecutter template for projects in Python. Your analysis doesn't have to be in Python, but the template does provide some Python boilerplate that you'd want to remove (in the src folder for example, and the Sphinx documentation skeleton in docs).

Requirements

• Python 2.7 or 3.5
• cookiecutter Python package >= 1.4.0: pip install cookiecutter

Starting a new project

Starting a new project is as easy as running this command at the command line. No need to create a directory first, the cookiecutter will do it for you.

cookiecutter https://github.com/drivendata/cookiecutter-data-science

Example

<script type="text/javascript" src="https://asciinema.org/a/9bgl5qh17wlop4xyxu9n9wr02.js" id="asciicast-9bgl5qh17wlop4xyxu9n9wr02" async></script>

Directory structure

├── LICENSE
├── Makefile           <- Makefile with commands like `make data` or `make train`
├── README.md          <- The top-level README for developers using this project.
├── data
│   ├── external       <- Data from third party sources.
│   ├── interim        <- Intermediate data that has been transformed.
│   ├── processed      <- The final, canonical data sets for modeling.
│   └── raw            <- The original, immutable data dump.
│
├── docs               <- A default Sphinx project; see sphinx-doc.org for details
│
├── models             <- Trained and serialized models, model predictions, or model summaries
│
├── notebooks          <- Jupyter notebooks. Naming convention is a number (for ordering),
│                         the creator's initials, and a short `-` delimited description, e.g.
│                         `1.0-jqp-initial-data-exploration`.
│
├── references         <- Data dictionaries, manuals, and all other explanatory materials.
│
├── reports            <- Generated analysis as HTML, PDF, LaTeX, etc.
│   └── figures        <- Generated graphics and figures to be used in reporting
│
├── requirements.txt   <- The requirements file for reproducing the analysis environment, e.g.
│                         generated with `pip freeze > requirements.txt`
│
├── src                <- Source code for use in this project.
│   ├── __init__.py    <- Makes src a Python module
│   │
│   ├── data           <- Scripts to download or generate data
│   │   └── make_dataset.py
│   │
│   ├── features       <- Scripts to turn raw data into features for modeling
│   │   └── build_features.py
│   │
│   ├── models         <- Scripts to train models and then use trained models to make
│   │   │                 predictions
│   │   ├── predict_model.py
│   │   └── train_model.py
│   │
│   └── visualization  <- Scripts to create exploratory and results oriented visualizations
│       └── visualize.py
│
└── tox.ini            <- tox file with settings for running tox; see tox.testrun.org

Opinions

There are some opinions implicit in the project structure that have grown out of our experience with what works and what doesn't when collaborating on data science projects. Some of the opinions are about workflows, and some of the opinions are about tools that make life easier. Here are some of the beliefs which this project is built on—if you've got thoughts, please contribute or share them.

Data is immutable

Don't ever edit your raw data, especially not manually, and especially not in Excel. Don't overwrite your raw data. Don't save multiple versions of the raw data. Treat the data (and its format) as immutable. The code you write should move the raw data through a pipeline to your final analysis. You shouldn't have to run all of the steps every time you want to make a new figure (see Analysis is a DAG), but anyone should be able to reproduce the final products with only the code in src and the data in data/raw.

Also, if data is immutable, it doesn't need source control in the same way that code does. Therefore, by default, the data folder is included in the .gitignore file. If you have a small amount of data that rarely changes, you may want to include the data in the repository. Github currently warns if files are over 50MB and rejects files over 100MB. Some other options for storing/syncing large data include AWS S3 with a syncing tool (e.g., s3cmd), Git Large File Storage, Git Annex, and dat. Currently by default, we ask for an S3 bucket and use s3cmd to sync data in the data folder with the server.

Notebooks are for exploration and communication

Notebook packages like the Jupyter notebook, Beaker notebook, Zeppelin, and other literate programming tools are very effective for exploratory data analysis. However, these tools can be less effective for reproducing an analysis. When we use notebooks in our work, we often subdivide the notebooks folder. For example, notebooks/exploratory contains initial explorations, whereas notebooks/reports is more polished work that can be exported as html to the reports directory.

Since notebooks are challenging objects for source control (e.g., diffs of the json are often not human-readable and merging is near impossible), we recommended not collaborating directly with others on Jupyter notebooks. There are two steps we recommend for using notebooks effectively:

1. Follow a naming convention that shows the owner and the order the analysis was done in. We use the format <step>-<ghuser>-<description>.ipynb (e.g., 0.3-bull-visualize-distributions.ipynb).

2. Refactor the good parts. Don't write code to do the same task in multiple notebooks. If it's a data preprocessing task, put it in the pipeline at src/data/make_dataset.py and load data from data/interim. If it's useful utility code, refactor it to src and import it into notebooks with a cell like the following. If updating the system path is icky to you, we'd recommend making a Python package (there is a cookiecutter for that as well) and installing that as an editable package with pip install -e.

# Load the "autoreload" extension
%load_ext autoreload

# always reload modules marked with "%aimport"
%autoreload 1

import os
import sys

# add the 'src' directory as one where we can import modules
src_dir = os.path.join(os.getcwd(), os.pardir, 'src')
sys.path.append(src_dir)

# import my method from the source code
%aimport preprocess.build_features
from preprocess.build_features import remove_invalid_data

Analysis is a DAG

Often in an analysis you have long-running steps that preprocesses data or trains models. If these steps have been run already (and you have stored the output somewhere like the data/interim directory), you don't want to wait to rerun them every time. We prefer make for managing steps that depend on each other, especially the long-running ones. Make is a common tool on unix platforms (and is available for Windows). Following the make documentation, Makefile conventions, and portability guide will help ensure your Makefiles work effectively across systems. Here are some examples to get started. A number of data folks use make as their tool of choice, including Mike Bostock.

There are other tools for managing DAGs that are written in Python instead of a DSL (e.g., Paver, Luigi, Airflow, Snakemake, Ruffus, or Joblib). Feel free to use these if they are more appropriate for your analysis.

Build from the environment up

The first step in reproducing an analysis is always reproducing the computational environment it was run in. You need the same tools, the same libraries, and the same versions to make everything play nicely together.

One effective approach to this is use virtualenv (we recommend virtualenvwrapper for managing virtualenvs). By listing all of your requirements in the repository (we include a requirements.txt file) you can easily track the packages needed to recreate the analysis. Here is a good workflow:

1. Run mkvirtualenv when creating a new project
2. pip install the packages that your analysis needs
3. Run pip freeze > requirements.txt to pin the exact package versions used to recreate the analysis
4. If you find you need to install another package, run pip freeze > requirements.txt again and commit the changes to version control.

If you have more complex requirements for recreating your environment, consider a virtual machine based approach such as Docker or Vagrant. Both of these tools use text-based formats (Dockerfile and Vagrantfile respectively) you can easily add to source control to describe how to create a virtual machine with the requirements you need.

Keep secrets out of version control

You really don't want to leak your AWS secret key or Postgres username and password on Github. Enough said, mostly — see the Twelve Factor App principles on this point. We generally use a .env file that, thanks to the .gitignore, never makes it into the repository (secrets should be shared via other means with contributors). The .env file defines secrets as environment variables, and is read in automatically by a package like dotenv in Python.

Contributing

The Cookiecutter Data Science project is opinionated, but not afraid to be wrong. Best practices change, tools evolve, and lessons are learned. The goal of this project is to make it easier to start, structure, and share an analysis. Pull requests and filing issues is encouraged. We'd love to hear what works for you, and what doesn't.

If you use the Cookiecutter Data Science project, link back to this page or give us a holler and let us know!

Links to related projects and references

Project structure and reproducibility is talked about more in the R research community. Here are some projects and blog posts if you're working in R that may help you out.

• Project Template - An R data analysis template
• "Designing projects" on Nice R Code
• "My research workflow" on Carlboettifer.info
• "A Quick Guide to Organizing Computational Biology Projects" in PLOS Computational Biology

Finally, a huge thanks to the Cookiecutter project (github), which is helping us all spend less time thinking about and writing boilerplate and more time getting things done.