latin-america-financial-development-lab

Data Recovery Audit

Last updated: 2026-07-08

Reconstruction update: This audit remains the evidence that the exact legacy processed files were absent from the repository and Git history. After this audit, a public-data reconstruction was completed. See docs/DATA_RECONSTRUCTION_LOG.md, data/processed/PanelCompleto.reconstructed.csv, data/processed/CreditType.reconstructed.csv, and data/processed/EconomicSector.reconstructed.csv.

Bottom Line

An exhaustive repository and Git-history audit found no recoverable analytical data files in latin-america-financial-development-lab. The repository contains legacy R scripts, TeX files, PDF procedure notes, and generated diagnostic outputs, but no committed raw or cleaned panels such as CreditType.final, EconomicSector.final, PanelCompleto.final, IPC, or country backup files.

Evidence Reviewed

  1. Current repository tree, including hidden files and generated outputs.
  2. Git history across all commits, branches, and tags.
  3. Script I/O references in all legacy and new R scripts.
  4. Targeted workspace search under the local GitHub workspace for expected dataset names.
  5. Procedure PDFs describing the legacy construction process.

Current Repository Findings

The only current .csv files are generated diagnostics and status outputs. No original .xlsx, .xls, .rds, .RData, .parquet, .feather, .fst, .dta, .zip, or country backup files were found in the repository.

Git History Findings

The repository history contains three commits:

No historical commit contains raw or cleaned data files.

Workspace Search Findings

A targeted search for CreditType, EconomicSector, PanelCompleto, IPC, old_tc, FinancialSystemBolivia, Economic Sector, and Credit Type did not find recoverable analytical datasets outside the repository. Matches were scripts, TeX/PDF files, or unrelated PortableGit documentation.

Required Data Files

The minimum files needed for full empirical reconstruction are:

Reconstruction Potential

The legacy scripts contain enough procedural logic to reconstruct the datasets if the source files are restored or if web downloads are rerun in a complete R environment. However, several countries require manual downloads or source-specific transformations, so full reconstruction is not guaranteed to be automatic without source access and package availability.

Decision

No substantive empirical results were generated because the required source data are absent. All current tables, figures, dashboards, and reports are diagnostic or status outputs, not research findings.

Artifacts Created