/data_parsers

Primary LanguageJupyter NotebookMIT LicenseMIT

Table of contents

Quick start

install anaconda or miniconda
conda create -n <project_name> python=3.9
pip install poetry
conda activate <project_name>
cd /path/to/this/repo
poetry install

Dicom

Tag based dicom file converter. Tags are not defined yet. Not used yet.

Excel

Nice to know

  • path_master.py -> holds src and export folder definitions
  • data in test folder -> healthy (myocarditis negative, not necessary really healthy)
  • data in train folder -> myocarditis positive

1. Pre-processing (to create basic data structure)

  • workbook_2_sheets.py -> extract sheets from workbook and save as separate files
  • sheets_2_tables.py -> extract tables from sheets and save as separate files
  • cleaner.py -> clean up tables and save in a new folder
  • checks.py -> check if all tables complete and save in a new folder

2. Refinement (more specific data arrangement for faster plotting)

  • calculate_accelerations.py -> calculate accelerations from raw data and save in a new folder
  • table_condenser.py -> focus on specific data and save in a same folder as acceleration results
  • table_merger.py -> merge tables and save in a new folder

3. Analyze (ce plots)

  • use jupyter notebook (load the data into the RAM for faster plotting iterations)