A working library of open-source tools I've built for research, evaluation, and data workflows — mostly in Stata and Python. Search or pick a problem below, then click a name for the source on GitHub, or the SSC tag to install from the SSC archive. Full index at github.com/ericabooth.
Publish interactive HTML/JS pages — charts, dashboards, tables, and maps — straight from a Stata do-file.
Call web APIs, read/write Google Sheets, work off Google Drive, pull in R data, query an LLM, and convert or edit any file — without leaving Stata.
Scrapers and downloaders for Texas education, health price-transparency, and Stata package archives.
Match and validate records, combine and reshape files, rebuild variables, generate codebooks, track data lineage, and de-identify data for safe public release.
Estimation and inference tools for evaluation work — matching, shrinkage, multilevel R-squared, simulation, conformal and nonparametric methods — plus applied research and replication materials.
Scaffold and back up whole analysis projects, install what a do-file needs, generate timesheets, auto-build multicam video timelines, and convert documents.
A draft book on applied program evaluation in Stata, an NSF-funded evaluation workshop, and a tool for building interactive SMCL help files and slides.
One screenshot from each of ten packages in the index above; each panel links to the source and, where there is one, a live demo.
Run mode executes the do-files with instrumentation around every join, so the map reports what matched, what share of each side took part, and which step removed the rows.
draw, export(mermaid) writes plain text naming the boxes and arrows. Paste that text into a README, a wiki or Quarto, and those pages render the diagram from it; to update the picture, re-export the text.
A column that is 52% missing is either a question people refused or a question they were never shown. surveymap counts refusals and skips separately for every item, and draws the path through the questionnaire that the answers imply.
Pick a state and every tile and bar updates for it, against a reference unit you choose. The build writes one HTML file that opens offline, with per-panel CSV and a save-as-PDF button.
A sparkta2 choropleth embedded as a panel in a dashboardbuilder page, with Jenks breaks, hover detail, zoom and a reset control. A Texas 2036 tool.
A searchable, sortable table and a histogram that redraws as you move the inputs. You write Stata; statashiny writes the HTML, CSS and JavaScript into one file you can email.
One Stata do-file wrote every chart on the demo page into a self-contained HTML file, tooltips and export menu included.
You keep writing Stata, and short commands compile to Bootstrap 5 pages with a navbar, in-page contents, collapsible code panels, captioned figures and embedded dashboards.
Each table becomes a high-resolution image, and the script then compiles the rewritten document and checks page by page that the layout still matches the original.
Fun experiments and personal Stata scripts that live in my adopath but aren't packaged on GitHub or SSC. Browse them in a shared Google Drive folder; in Stata, type which cmd to see any command's syntax.
graph bar/hbar with its mean as a % label (and n), honoring by() and over().
logo option reprints the Stata banner.obs, observations) whose share of missing values passes a threshold..tex file to a PDF straight from Stata (macOS or Windows).Source for everything is at
github.com/ericabooth →;
all my Stata modules are also on RePEc / SSC →.
A SSC tag means the package installs with ssc install <name>.
Longer walkthroughs are in the mergemap and surveymap galleries and on the
sparkta2,
dashboardbuilder,
statashiny,
googlechart and
webdoc2 demo sites.