KaiquanMah/DSIP

{{DSIP.project}}

Inspired by Serj Smorod DS in Production lecture on project structure and best practices for data science.

Project Structure

• —app/: Helper scripts and utilities.
• —models/: Trained models for the current experiment.
• —archived_experiments/: Archived experiments and their outputs.
• —data/: Input datasets and preprocessed data.
• —results/: Outputs like predictions, charts, and analysis results.
• —notebooks/: Jupyter notebooks for exploration and experimentation.
• —tests/: Unit tests to ensure code quality.

Core scripts include:

• —preprocess.py: Handles data preprocessing tasks.
• —train.py: A script to train machine learning models.
• —predict.py: Generates predictions using trained models.
• —result.py: Analyzes results and generates metrics/charts.
• —tasks.py: Automates workflows using invoke.