Skip to content

enemduR v0.1.0 — First stable portfolio release

Latest

Choose a tag to compare

@yerovi84 yerovi84 released this 02 Jun 19:05
837c0a6

enemduR v0.1.0 — First stable portfolio release

enemduR v0.1.0 is the first stable portfolio release of the package.

It provides reproducible analytical infrastructure for Ecuador's ENEMDU microdata workflows, including reading, validation, metadata registries, survey-design-aware estimation, representativity assessment, labor indicators, income, quintiles, poverty, NBI, IPM/TPM, benchmark comparison workflows, pkgdown documentation, and portfolio vignettes.

This release positions enemduR as an analytical engine. It is not a dashboard, not a CRAN release, not an official statistical production system, and not official institutional validation.

Installation

remotes::install_github("yerovi84/enemduR@v0.1.0")
library(enemduR)
packageVersion("enemduR")

Expected version:

[1] '0.1.0'
Validation summary

Validated locally with:

git diff --check
devtools::load_all(reset = TRUE)
devtools::test()
pkgdown::build_site()
devtools::check()
installation from GitHub tag yerovi84/enemduR@v0.1.0

Reported results:

devtools::test(): 1438 passed, 0 failed, 0 warnings, 0 skipped.
pkgdown::build_site(): passed.
devtools::check(): 0 errors, 0 warnings, 0 notes.
installation from tag: passed.
installed package version: 0.1.0.
Scope

This release includes infrastructure for:

reading .sav, .dta, and .csv files;
metadata registries;
survey-design-aware estimators;
representativity and precision assessment;
labor indicators based on condact;
income, quintiles, poverty, and extreme poverty;
NBI;
IPM/TPM;
benchmark comparison workflows;
pkgdown documentation;
portfolio vignettes.
Methodological guardrails

This release does not claim official institutional validation.

It preserves explicit methodological boundaries, including:

no reconstruction of condact from raw questionnaire variables;
no poverty classification without explicit, auditable poverty lines;
no global sentinel-value recoding;
no design-domain prefiltering before estimation;
no confusion between weighted population counts and effective sample size.
Notes

This is a GitHub-based stable portfolio release. Future work may include broader production hardening, CRAN readiness, historical harmonization, raw labor-status reconstruction if methodologically authorized, and downstream Quarto/dashboard products that consume the analytical engine.