Coding & Computational Projects
Repositories with a strong coding and computational focus, ranging from econometric research pipelines to data infrastructure and computational economics coursework. All projects are open-source on GitHub.
MapleStats MCP: One MCP Server for Canadian Open Data
Canadian public data is spread across Statistics Canada tables, Census products, public use microdata files, the Bank of Canada, CMHC, federal agencies, and dozens of federal, provincial, territorial, and municipal open-data portals, each with its own API and conventions. MapleStats MCP is a Model Context Protocol server, built in Python, that gives AI assistants typed, queryable access to all of it through a single interface, in English and French. Every result carries its source, so figures can be traced back to the agency that published them. It installs from PyPI with a single uvx maplestats-mcp command and works with Claude, Codex, Cursor, and any other MCP client.
EcuDataMCP: An MCP Server for Ecuador’s Open Government Data
Ecuador’s public data is abundant but scattered — household and labour surveys, the companies registry, procurement records, seismic feeds, regulatory texts — each living behind a different agency, a different API, and often Spanish-only documentation. EcuDataMCP is a Model Context Protocol server, built in Python, that collapses all of that into one interface: a single set of structured, queryable tools that let AI assistants search, filter, and analyze Ecuador’s open government data directly, without ever leaving the conversation. Under the hood it wires together CKAN’s open-data catalog, SERCOP’s public procurement records, the Instituto Geofísico’s real-time seismic and tsunami-risk feeds, the Supercías companies and financial registries, INEC’s survey and census catalog, and directories of government institutions and regulations, turning a dozen disconnected sources into one coherent, assistant-readable layer.
Do investment tax credits foster innovation? The AITC and patent applications
A two-way fixed effects difference-in-differences study of the Alberta Investor Tax Credit (AITC) and its effect on patent applications, implemented entirely in R. The repository includes raw data pipelines, cleaning scripts, econometric estimation, and a fully reproducible Quarto manuscript.
Sánchez Pazmiño, D. H. (2024). Do investment tax credits foster innovation? The Alberta Investor Tax Credit and patent applications. MA Economics Capstone Paper, Simon Fraser University.
Retrospectively sweaty? Temperature changes and presidential approval
An R-based research pipeline matching daily CPC Global Unified temperature raster data to AmericasBarometer survey microdata at the respondent level, using geocoded coordinates and date of interview. Includes data cleaning, spatial joins, and panel regression estimation.
Sánchez Pazmiño, D. H. (2024). Retrospectively sweaty? The effects of temperature changes on presidential approval. POL837 Issues in Comparative Politics Research Paper, Simon Fraser University.
ECON832: Computational Methods for Economics
Course repository for ECON832 at Simon Fraser University, covering computational methods used in economic research. Includes lecture notes, problem sets, and code examples implementing numerical and simulation-based techniques in Julia and R.
ecuador-temperature-noaa: Ecuador Climate Data from NOAA
A data processing pipeline to retrieve and transform daily climate data for Ecuador from NOAA’s gridded datasets. Raw NetCDF geospatial files are converted into structured CSV outputs at the canton level, covering maximum temperature, minimum temperature, and precipitation, making the data suitable for panel regressions and environmental or health research in Ecuador. Maintained under LIDE.
Ecuador data jobs scraping and labour-market analysis
An R-based data pipeline I built to collect and analyse public job postings for data and technology-adjacent roles in Ecuador. The scraper standardizes postings from Multitrabajos, Computrabajo, and Mipleo into a common schema, extracts job attributes and skill requirements, classifies postings into data job families, and produces repeatable cross-sections for labour-market analysis. A separate analysis repository accumulates scrape waves, deduplicates listings, and generates the figures and tables used for an El Quantificador article on the tools and technical skills demanded in Ecuador’s data labour market.