District-wise nighttime lights database for India (641 districts, 2012-2024) using VIIRS satellite data.

Replication Code for "Supporting Teacher Autonomy to Improve Education Outcomes: Experimental Evidence from Brazil"

BREAD Working Papers

This repository contains the replication files for the paper "Do Private Consultants Promote Savings and Investments in Rural Mozambique?"

Here you can have access to our group Master Thesis 2022 code by Stata.

Class project

Data science & econometrics project exploring whether micro-finance institutions prioritise borrower need, and whether repeated borrowing promotes financial inclusion. We use LLM-labelled loan narratives and ML to prepare over 1 million Kiva loans for a fixed effects regression analysis.

Donor-ready economic evaluation framework and KPI system for the Awaken Foundation’s cacao value-chain program in Ecuador, integrating applied econometrics and impact-measurement methods.

Replication and extension of Besley & Burgess (2004) using modern DiD, event-study diagnostics, heterogeneity, and robustness checks in R.

Research deliverable developed during voluntary collaboration with Youth Advocacy Africa (YAA), Nairobi — Focusing on aligning youth-led initiatives with global funding priorities and SDG-linked donor frameworks (2025–2026).

Replication material to reproduce the Corrigendum to "The Long-run Impact of Bombing Vietnam"

Code sample repository to showcase abilities in various programming languages.

A structured repository of Python scripts and Jupyter notebooks for development sector data workflows — including public health, gender equity, women's economic empowerment (WEE), education, and MEL (Monitoring, Evaluation, and Learning). Includes plug-and-play templates, sample data, test coverage, and Colab-ready execution.

"Causal Machine Learning for Cost-Effective Allocation of Electricity Aid" thesis for my Masters in Management and Digital Technologies at Ludwig-Maximillian Univeristy, Munich.

Raship with Alessandro Sforza and Tommaso Sonno. Codes for manipulation of IO (Input-Output) table.

Luna's Personal Website

End-to-End Python implementation of Azcue et al.'s (2025) stochastic optimal control framework for social protection policy design. Solves PDMP-based Hamilton-Jacobi-Bellman equations using analytical closed-form solutions and Monte Carlo simulation to minimize government intervention costs (through the use of cash transfers and microinsurance).

End-to-End Python replication of "Who Connects Global Aid?" (McCarthy et al., 2025). Uses bipartite network projection and Node2Vec embedding for analyzing 10M+ IATI transactions. Maps structural influence in global aid using HITS centrality, betweenness analysis, and UMAP visualization.

End-to-End Python implementation of Dávila-Fernández & Sordi's (2025) methodology for FX-constrained growth modeling (in emerging markets). Features Bayesian state-space estimation via Gibbs sampling with FFBS algorithm, heterogeneous agent simulation (fundamentalists/chartists), and nonlinear dynamics analysis.