Skip to content

REPRODUCIBILITY.md

Latest commit

 

History

History
28 lines (22 loc) · 1009 Bytes

REPRODUCIBILITY.md

File metadata and controls

28 lines (22 loc) · 1009 Bytes
Raw

Reproducibility Notes

Local CLI

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
bash scripts/run_pipeline.sh

The default pipeline creates demo numeric features if data/processed/features.csv is absent. Those demo results only prove that the code path runs; they are not scientific results.

Google Colab

  1. Upload or clone this repository in Colab.
  2. Install requirements with pip install -r requirements.txt.
  3. Store logs and checkpoints in Google Drive if the runtime may disconnect.
  4. Do not upload identifiable patient data unless your institution has approved that exact cloud workflow.

Required Real Inputs

For thesis experiments, replace demo features with a real CSV at data/processed/features.csv:

sample_id,feature_1,feature_2,...,label
case_001,0.23,1.91,...,0
case_002,0.74,0.88,...,1

Keep train/validation/test splits patient-level. Do not place slices or images from the same patient in both training and testing.