Skip to contents

Changelog

Source: NEWS.md

QuickExplore 0.1.1

Bug fixes

  • Code Generator: closed the missing parenthesis on the strat_levels line emitted for stratified cross-tabs. The generated .R script now parses cleanly (#BUG-03).
  • Code Generator: normalise Windows back-slashes in file paths so the generated .R script is parseable on every platform (#BUG-04).
  • Code Generator: emit dplyr::select(dplyr::all_of(c(...))) instead of bare column names, so non-syntactic names (spaces, hyphens, leading digits) survive the round-trip (#BUG-06).
  • read_dataset(): preserve haven label / format.sas attributes and the haven_labelled class when converting blank SAS strings to NA (#BUG-05). get_variable_info() once again populates Label and Format for SAS inputs.
  • read_dataset(): reject .rds files whose top-level object is not a data.frame, with a clear error message instead of cryptic downstream failures (#BUG-20).
  • format_file_size(): extended units to TB and PB and added a guard for negative sizes (#BUG-07).
  • get_variable_info(): N_Unique now counts distinct non-missing values, matching skimr and DataExplorer (#BUG-08). Missing counts continue to be reported separately in Missing_Count.
  • compute_numeric_summary() / compute_categorical_summary(): emit a clear error naming the offending variable when vars includes a name that is not in df, instead of base R’s undefined columns selected (#BUG-09).
  • compute_numeric_summary(): now returns Q1 and Q3 quartile columns alongside the existing summary statistics (#BUG-13).
  • compute_crosstab(): validate row_var, col_var, and strat_var up-front and fail with an actionable error if any is missing or duplicated (#BUG-15).
  • compute_crosstab(): warn when the source data carries values literally equal to Total or (Missing), which would otherwise be indistinguishable from synthetic totals / NA-relabel cells (#BUG-14).
  • Converter: the CSV preview now honours the user’s delimiter and header-row toggles, matching what the download handler writes (#BUG-11).
  • Converter: rewrote XPT variable-name shortening so a make.unique step is no longer followed by truncation, avoiding duplicate 8-character names; if any name is shortened, a _name_map.csv is emitted alongside the .xpt so analysts can trace renames (#BUG-10).
  • Dataset Browser: removing the currently active library now resets selected_dataset and loaded_data so the Data Viewer, Convert, and Code tabs stop showing stale content (#BUG-16).

Documentation

  • Removed contradictory @keywords internal on code_generator_server, which had been exported simultaneously (#BUG-17).
  • Documented compute_crosstab(), the Cross-tabulation panel, the Code Generator tab, and the SAS blank-to-NA coercion that were already shipped in 0.1.0 (#BUG-18).

Tests

  • Added tests/testthat/test_quickexplore_suite.R: testthat edition-3 suite with eight thematic sections (~40 test cases) covering file-format dispatch, metadata, numeric/categorical summaries, cross-tabulation, converter round-trip, code-generator parseability, and data-type coverage.
  • Added tests/testthat/test_quickexplore_runner.Rmd: an R-Markdown runner that knits to an HTML report with pass/fail tallies and inline bug-regression demonstrations.

QuickExplore 0.1.0

CRAN release: 2026-04-21

Initial Release

  • run_app() launches the Dataset Explorer Shiny application.
  • Browse multiple directory-based libraries simultaneously.
  • Load datasets in SAS (.sas7bdat, .xpt), CSV, and R (.rds) formats.
  • Interactive data viewer with column-level filtering via DT.
  • Explore panel with dplyr-expression filtering and quick-filter badges.
  • Variable Explorer with type-coloured metadata and top-value frequency tables.
  • Summary panel with automatic numeric and categorical statistics, plus a missing-value heat bar table.
  • Converter panel exports to .rds, .xlsx, .csv, .json, and SAS V5 transport (.xpt).
  • Exported utility functions: read_dataset(), list_datasets(), get_variable_info(), get_dataset_metadata(), format_file_size(), compute_numeric_summary(), compute_categorical_summary().