LoRa HAR Evaluation

Lightweight Python pipeline for evaluating human activity recognition from WiMOD LoRa logs.

What it does

• merges datastudio*.log and datarlt*.log
• extracts windowed RSSI/SNR/PER features
• runs either multiclass or binary_anomaly evaluation
• saves predictions and optional plots

Project structure

• main.py – main evaluation runner
• config.py – segment schedule, window size, mode
• postprocess.py – multiclass postprocessing
• util/ – parsing, features, metrics, models, plots
• run.bat – runs all postprocess variants

Install

pip install -r requirements.txt