Household mobility responses to weather extremes in Kyrgyzstan

This repository contains the data-processing pipeline, spatial econometric estimation code, robustness analyses, and figure-generation scripts for the study: "Household mobility responses to weather extremes in Kyrgyzstan" The repository reproduces: panel-data construction, spatial weights generation, multinomial logit estimations with elevation interactions, robustness and diagnostic checks, manuscript figures and source tables. Repository structure: data/raw/ contains the required raw input datasets. output/model_outputs/ contains estimated model outputs. output/figures/ contains manuscript-ready figures. output/source_tables/ contains source data underlying figures. output/robustness/ contains robustness and diagnostic outputs. output/W/ contains spatial weights matrices used in estimation. Main scripts: mobility_core.py — core data preparation and estimation pipeline robustness_checks.py — robustness and post-estimation diagnostics visuals.py — figure and map generation mobility_pipeline_main.py — master execution pipeline The code was developed in Python using pandas, geopandas, matplotlib, numpy, and spatial econometric utilities. All paths are configured relative to the repository root to support reproducible execution in archival and computational environments such as Zenodo and Code Ocean.