Urban Walk Dataset – Lisbon 2024

This dataset contains detailed physiological, environmental, spatial, and perceptual data collected during 2,207 pedestrian trips made by 90 participants in Lisbon, Portugal, between August and December 2024. Each row corresponds to a unique walking route and includes: Self-reported affective states (well-being, tiredness, agitation, environmental perception) Physiological signals: heart rate variability (HRV), electrodermal activity (EDA), and skin temperature Environmental data: NDVI (vegetation index), noise levels, temperature, precipitation, and route slope Visual features: segmented urban elements detected via Mapillary street images Spatial context: Points of Interest (POIs) extracted from OpenStreetMap The dataset is accompanied by a full variable description table (Metadata.csv) and the Python/JavaScript scripts used for data collection, processing, and statistical modeling of the three main hypotheses explored in the study. 📁 Included Files: Data_per_route.csv: Main dataset (1 row per trip) Metadata.csv: Variable descriptions (type, unit, possible values) 1.APMA_data.py: Weather data processing 2.NDVI_LISBON_GEE.js: NDVI extraction via Google Earth Engine 3.Mapillary_collect.py: Object detection via Mapillary API 4.Mapillary_segmetation.py: Image segmentation aggregation 5.Osm_data.py: Extraction of POIs from OpenStreetMap 6.Model_H1.py, 7.Model_H2.py, 8.Model_H3.py: Statistical modeling scripts