Changelog

Source: NEWS.md

datacleanr 1.0.4

  • improved plotting performance of large, “narrowly distributed” distributed data, such as a 1:1 scatter or a high resolution time series (i.e., little variation in either x or y dimension). The low performance seems to come from changes made in how hover events are handled in plotly in newer versions (2.x.x). See this issue, which is still unresolved at 06 June, 2022. The fix uses the default hovermode = "closest", but sets hoverdistance = 200, which should ensure a better responsiveness in most cases. The increased hoverdistance called for a disabling of hover labels to prevent confusion. Note, the increased hoverdistance may cause some unintended consequences when clickin into plotting ‘white space’, as the click snaps to the closest point within 200 pixels.

datacleanr 1.0.32021-11-18

  • fixed filter_scoped() test to conform with dplyr 1.0.8 release (dropped dplyr error class)

datacleanr 1.0.22021-11-03

  • removed navbarPageWithInputs, replaced with shiny::navbarpage() and bslib::nav_item(), as recommended
  • adjusted button placement (close/exit) with css styles
  • plotly no longer serves Plotly.d3, which was used to select individual traces on the plot in the viz tab; it now requires additional import. This is currently done via a D3 script dependency from a CDN, which requires an internet connection (checked by can_internet()

datacleanr 1.0.12021-02-10

  • dcr_checks:
    • Added support for fs_path-type filepaths (from fs package) when launching dcr_app() with a path to an*.Rds file.
    • Added check for existing .annotation column, which produces useful error when present.
    • Added check for data sets above 1.5 million observations (rows) - this threshold is at the limit of modern browsers for visualizing points with D3/plotly.
  • Updated minimum versions for R and dependencies
  • Changed summarytools::view() to print() for data overview (dropping dependency on X11 / XQuartz)
  • Separated setting grouping structure and enabling filtering, plotting, and extracting from data frame summary. This saves significant time for larger data sets. Help links and panel texts were adjusted accordingly.

datacleanr 1.0.02020-11-02

  • First major version with features for exploration, filtering, visualizing and cleaning/annotating data in a reproducible manner.
  • Added a NEWS.md file to track changes to the package.