MedDataKit 0.0.12

Medical Public Data Kit

MediDataKit

Design

1. Downloader: download data from specific sources e.g. UCIML, Local Directory

   • Functionalities:
     • download data files with provided URLs, Paths, and other information
     • save downloaded files, check if the file exists before downloading
   • Information produced:
     • raw_data_files: original downloaded data files (file name and paths, number of files, file size, etc.)

2. DataLoader: load data into memory

   • Functionalities:
     • load downloaded data into memory

raw_data_files (original downloaded data files) -> raw_data (centralized data, federated data) -> ML-ready data (data ready for performing ML tasks)

Included Datasets

clinical dataset

• ZigongHeartFailureDataset
• MIMIC2IACDataset
• MIComplicationsDataset
• DiabeticHospitalDataset
• ARI2Dataset
• SupportDataset
• CIBMTRHCTSurvivalDataset
• CrashDataset
• RHCDataset
• HCCSurvivalDataset
• ZAlizadehsaniDataset
• NasarianCADDataset

repo dataset

• ArrhythmiaDataset
• ColposcopyDataset
• SPECTFDataset
• BreastCancerWisconsinDataset
• DermatologyDataset
• BoneTransplantDataset
• ParkinsonsDataset
• CervicalRiskDataset
• BacteremiaDataset
• FetalCTGDataset

bioinfo dataset

• CodonUsageDataset
• GENE3494Dataset

APP

shiny run --reload --launch-browser shiny-app/app.py