LoginSign Up
← Back to all sparks
D

daiquiri

ANALYTICS
Velocity0.0

Data quality reporting for temporal datasets. Contribute to ropensci/daiquiri development by creating an account on GitHub.

github.com/ropensci/daiquiri ↗

A data-quality report generator that finished its API rewrite and has been coasting on small features since.

data-qualityr-packagereportingropenscitime-series
◆Current state
daiquiri turns a raw clinical or administrative dataset into an HTML report of time-series data-quality plots, driven by a field-type specification the user writes. The public API settled in 2022 after a wholesale rename for rOpenSci acceptance, and releases since then have added specification conveniences rather than new report content. The 1.2.0 release is the first in nearly two years.
◆Where it's heading
Development has shifted from restructuring the interface to lowering the cost of using it — field_types_advanced() lets users name only the columns they care about and default the rest, which is the kind of change that matters when a dataset has hundreds of fields. Plot rendering is getting incremental attention (heatmap scaling) rather than new visualisation types. Cadence is roughly annual.
◆Prediction
Expect the next release to continue trimming specification boilerplate for wide datasets rather than adding report sections; the entries give no indication of a new plot type or output format in progress.

◆Recent moves

  1. 1y ago

    field_types_advanced() lets specs name only some columns

    The first release in almost two years, and it targets the specification step rather than the report. Naming a subset of columns and defaulting the rest removes the main friction in pointing daiquiri at a wide dataset.

    View source ↗
  2. 3y ago

    ft_strata() splits reports by a column's values

    Stratified reporting is the most substantive report-content addition in the window, letting one run produce per-group aggregate and field plots. The rest of the release is post-rename cleanup — missing-value strings, character handling, and rmarkdown chunk-label conflicts.

    View source ↗
  3. 3y ago

    Column-order and integer-column validation fixes

    Bug fixes cleaning up after the 1.0.0 rename, including a chunk-label regression introduced one release earlier. Nothing here changes what a report contains.

    View source ↗
  4. 3y ago

    Report intermediates write to tempdir(), not the library

    A packaging fix so reports render when the library directory is read-only. Invisible to anyone whose installation already worked.

    View source ↗
  5. 3y ago

    First CRAN release

    Reaching CRAN is the distribution milestone that follows the 1.0.0 rename — the renamed API is what got accepted. The only code change is dropping a deprecated ggplot2 call.

    View source ↗
  6. 3y ago

    Public API renamed wholesale for rOpenSci acceptance

    ⚡ SPARK

    This is the release the current API dates from: create_report(), fieldtypes(), log_initialise() and their siblings were all renamed at once as the price of rOpenSci review. Everything after it is refinement of this surface.

    View source ↗