LoginSign Up
← Back to all sparks
B

bulkreadr

INFRA · APIS
Velocity0.0

The Ultimate Tool for Reading Data in Bulk

github.com/gbganalyst/bulkreadr ↗

A bulk file reader became a labelled-survey-data toolkit, then went quiet

data-importsurvey-datalabelled-dataspss-statadependency-reduction
◆Current state
bulkreadr started as a way to read many files at once and turned into tooling for labelled survey data: SPSS and Stata importers that convert labelled variables to factors, generate_dictionary() for building data dictionaries, look_for() for searching variable descriptions, and imputation helpers. The most recent release does the opposite of adding — it pulls inspect_na() in-house to drop an external dependency.
◆Where it's heading
Growth came in a burst across 2023, slowed to one release a year, and has now turned inward. The 2023 cadence added a format or a labelled-data function every few weeks; 2025 added a single Excel-to-CSV exporter; 2026 removed a dependency. The GitHub notes are cumulative — each release restates every prior version's changelog — which makes the feed look busier than the work is.
◆Prediction
With inspectdf gone, the remaining Suggests-level dependencies are the obvious next targets for the same treatment. Nothing in these entries points to a new file format or a return to the 2023 pace.

◆Recent moves

  1. 5mo ago

    inspect_na() brought in-house to drop the inspectdf dependency

    Missingness summaries are reimplemented internally so the package stops importing inspectdf. Users see the same function; the change is entirely in the dependency graph.

    View source ↗
  2. 1y ago

    Every Excel sheet exported to its own CSV file

    write_excel_sheets_to_csv() adds an export direction to a package built for reading, splitting a workbook into one CSV per sheet with consistent naming. Note the body announces itself as version 1.1.2 while the tag reads 1.2.0.

    View source ↗
  3. 2y ago

    Six imputation strategies for fill_missing_values()

    Imputation moves from mean-only to a menu — minimum, maximum, mean, median, harmonic and geometric mean — over the same data. More estimators rather than a new capability, plus the launch of the package website.

    View source ↗
  4. 2y ago

    Data dictionaries and keyword search over labelled variables

    generate_dictionary() builds a variable dictionary from a labelled dataset and look_for() searches names and descriptions by keyword. Both target the problem that arrives with SPSS and Stata files: hundreds of variables nobody can navigate.

    View source ↗
  5. 2y ago

    Stata .dta import alongside SPSS

    read_stata_data() gives Stata files the same labelled-to-factor treatment SPSS got one release earlier, and dependencies are trimmed. A second format on an established path rather than a new one.

    View source ↗
  6. 2y ago

    SPSS import converting labelled variables to factors

    ⚡ SPARK

    The turn in the package's history. read_spss_data() handles .sav and .zsav files and converts labelled variables into factors, which is what pulls bulkreadr out of generic file reading and into survey data — everything since, from Stata support to data dictionaries, follows from this.

    View source ↗