KNSI GOLEM Template Repository

Template repository for KNSI GOLEM python projects made from the cookiecutter data science template.

This repository's goal is to provide out-of-the-box boilerplate code with a clean project file structure. Additionally, the template includes a GitHub CI pipeline with pytest and flake8 checks implemented.

What is here?

Code and Files Structure

To ensure a transparent and easily understandable file structure for external users each module, from making plots to training models, is given its respective file in the src folder. A full description of the file structure is provided in the Project Organization section.

How to run code in repository?

First of all you need to install requirements

venv

Python's venv allows you to easily manage isolated environments for your projects, enabling you to work with specific module/library versions or even different Python versions without conflicting with the globally installed Python interpreter.

How to use

python3 -m venv .venv  # Create venv  
source .venv/bin/activate  # Activate venv  
pip install -r requirements.txt  # Install requirements to venv  

Then you need to create .env file in project directory that conatains password to soccernet data .env file example:

PASSWORD = "password_to_unzip_data"

Download dataset

python3 -m src.download_dataset # Download and unzip dataset

Extracting Frames from Videos

python3 -m src.extract_frames_snb

Usage:

python3 -m src.extract_frames \
    --sample_fps 5 --num_workers 3

Train the model

python3 -m src.modeling.train_tdeed_bas --model SoccerNetBall_baseline

SoccerNetBall_baseline is json file stored in modelconfig directory, you should start training twice - first on the "store_mode": "store" and after that "store_mode": "load."

dotenv

The dotenv library allows you to define environment constants or secrets, such as API keys, in a single place. It simplifies the management of environment variables by letting you configure them in a .env file.

The python-dotenv library has already been added to the requirements.

.env file example:

API_KEY = "KNSI_GOLEM_API_KEY"

python code example:

from dotenv import load_dotenv
import os

load_dotenv()

api_key = os.getenv("API_KEY")

Project Organization

├── 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               <- Project's docs
│
├── 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`.
│
├── 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.cfg          <- Configuration file for flake8 and pytest
│
└── src   <- Source code for use in this project.
    │
    ├── __init__.py             <- Makes src a Python module
    │
    ├── config.py               <- Store useful variables and configuration
    │
    ├── dataset.py              <- Scripts to download or generate data
    │
    ├── features.py             <- Code to create features for modeling
    │
    ├── modeling                
    │   ├── __init__.py 
    │   ├── predict.py          <- Code to run model inference with trained models          
    │   └── train.py            <- Code to train models
    │
    └── plots.py                <- Code to create visualizations