HONEYCOMB

just another minimalist data science skeleton

Project Organization

├── LICENSE
├── 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.
│   ├── features                 <- Features extracted from raw data using domain knowledge.
│   ├── processed                <- The final, canonical data sets for modeling (merge between interim and features).
│   └── 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
│
│
├── references                   <- Data dictionaries, manuals, and all other explanatory materials.
│
│
├── requirements.txt             <- The requirements file for reproducing the analysis environment, e.g.
│                                    generated with `pip freeze > requirements.txt`
│ 
├── notebooks                    <- Jupyter notebooks (Source code for use in this project and EDA)
│    │
│    │
│    └── visualization           <- Scripts to create exploratory and results oriented
│
└── src                          <- Main Python script (Source code for use in this project)
    │
    ├── __init__.py              <- Makes src a Python module
    │
    ├── data                     
    │   └── make_dataset         <- Scripts to download or generate data
    │   │         
    │   └── data_io              <- Functions to load and save data
    │   │         
    │   └── preprocess_tabular   <- Functions to preprocess raw tabular data
    │   │         
    │   └── preprocess_sound     <- Functions to preprocess sound or time series data
    │   │         
    │   └── preprocess_sound     <- Functions to preprocess text raw data
    │   │      
    │   └── preprocess_image     <- Functions to preprocess raw image data
    │
    │
    ├── features                 <- Scripts to turn raw data into features for modeling
    │   │         
    │   └── tabular_feature_engineer   <- Functions to make feature engineer from raw tabular data
    │   │         
    │   └── sound_feature_engineer     <- Functions to make feature engineer from raw data sound or time series data
    │   │         
    │   └── text_feature_engineer      <- Functions to make feature engineer from text data 
    │   │      
    │   └── build_features             <- main script from features eng. function are called 
    │
    │
    ├── models                   <- Scripts to train models and then use trained models to make predictions
    │   │                                               
    │   ├── run.sh               <- bash script where train and test experiment are setted up
    │   │ 
    │   └── train                <- main script to train model
    │   │ 
    │   └── inference            <- main script to inference from trained model   
    │   
    │   
    │   
    ├── fold                    <- Scripts to create folds from datasets 
    │   
    └── metric                  <- Scripts with many validation functions  

---

honeycomb

Honeycomb is a simple Machine learning framework with the aim to have a simple but useful starting point for data science projects and kaggle competitions. Some things have be borrow from other projects and have been adapted to it.