tqdm allows for progress bars when testing out code, good for determining length of runtime.

---

This article is to determine the best practices for project structure, environments & reproducibility, coding for reusability, testing, and collaborating.

Project Structure

defined by Cookie Cutter DS:

├── 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`
│
├── setup.py           <- Make this project pip installable with `pip install -e`
├── 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.readthedocs.io

Environment:

• Try to create the minimum reproducible environment so there is less dependencies so the project can be run locally by anyone.
  • docker/vagrant (for complex environments for virtual machines)
• 
• watermark allows people to know what version certain things are

• 

• to create isolated environments: conda env to create an environment on your machine

  conda create -n <name> python =3

  • venv, pyenv are alternatives
  • name environment as project root

• conda env list to see all environments

• include a requirements.txt to include dependancies where:

  

  • to install all requirements, go into the directory of this file and do pip install -r requirements.txt
  • pip freeze > requirements.txt to create

• so activate environment, install necessary dependencies, then deactivate when necessary

• create pip project:

  

• %debug in jupyter notebooks as a magic tool to debug code
  • use u to go up the error code, can use functions inside
  • use q to quit the debugger
• %prun to find computational inefficiencies for memory and runtime

Testing

• use pytest and assertions to see and test code
• use @pytest.fixture for static data referencing

–

More resources:

1. Video where notes are taken from
2. Cookie Cutter DS Structure
3. DSLP Structure
4. Labs