ayulockin / kagglerecipes

Helpful data preprocessing, training, and visualisation code and scripts for a range of Kaggle competitions, supported by Weights & Biases.

Home Page:https://ayulockin.github.io/kagglerecipes/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Kaggle Recipes

kaggle_wandb.png

What's New

Sept 3, 2021

  • Bug fix

Sept 2, 2021

  • Release version 0.0.3

  • Added support for multiprocessing to extract DICOM metadata for RSNA-MICCAI Brain Tumor Classification competition.

  • Bug fixes.

Aug 27, 2021

  • Released the library on PyPI.

  • Easily create voxel manipulated dataset for RSNA-MICCAI Brain Tumor Classification competition.

  • Extract dicom metadata.

  • Added utilities to log dataframe as tables and files/directory as artifacts.

  • Added utiities to log basic W&B charts (line, bar, and scatter).

Kaggle Competitions

Code & Scripts

RSNA-MICCAI Brain Tumor Radiogenomic Classification

  • LINK TO NOTEBOOK

Install

pip install kagglerecipes

Sample Datasets

We have also logged smaller subsets of Kaggle commpeition datasets local development and fast prototyping.

RSNA-MICCAI Brain Tumor Radiogenomic Classification

  • Download it manually from here.

  • Or download it using this code snippet.

    
    import wandb
    
    run = wandb.init()
    
    artifact = run.use_artifact('wandb_fc/rsna-miccai-brain/sample:v0', type='dataset')
    
    artifact_dir = artifact.download()
    
    

Credits

The code in this repository is a combination of many of the Kagglers' work that they have shared publicly as Kaggle kernels.

Note that this is not an official Weights and Biases product.

About

Helpful data preprocessing, training, and visualisation code and scripts for a range of Kaggle competitions, supported by Weights & Biases.

https://ayulockin.github.io/kagglerecipes/

License:Apache License 2.0


Languages

Language:Jupyter Notebook 80.7%Language:Python 18.8%Language:Makefile 0.4%