Skip to content

Releases: AnthonyRaborn/ShortForm

ShortForm 0.5.8

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 06 May 16:57

ShortForm v0.5.8

Bugfixes

  • Fixed issue in tabuShortForm for certain parallel workflows
  • Fixed bug in tabuShortForm when using multidimensional models

Updates

  • Removed dependencies on ggplot2, ggrepel, and tidyr (ACO and TS plot methods)
  • Added some additional unit tests

To Do

Note--these are not required for 0.5.8, but are on the roadmap for future updates.

  • add additional unit tests (maybe for 0.5.9)
  • refactor all major functions (0.6.0)
  • standardize function inputs, arguments, and outputs (0.7.0)

CRAN v0.5.7

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 04 May 02:06

ShortForm v0.5.7

Bugfixes

  • Fixed issue in ACO algorithm where the best model was not properly updating
  • Corrected use of ggplot within the ACO plot method due to depreciated ggplot2 functions
  • Fixed issue in SA algorithm where models including factor relationships or outcome variables were not specified correctly after the initial model

Updates

  • Modernized the README file
  • Fixed CI/CD badge since Travis CI no longer works for this project

To Do

Note--these are not required for 0.5.7, but are on the roadmap for future updates.

  • add additional unit tests (maybe for 0.5.8)
  • refactor all major functions (0.6.0)
  • standardize function inputs, arguments, and outputs (0.7.0)

CRAN v0.5.6

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 03 Dec 01:06

ShortForm v0.5.6

  • Updates focused on {antcolony.lavaan}.
    • lavaan.model.specs now defaults to the default arugments. If a user removes the arguments, they will be provided for them. If a user overwrites a specific argument (e.g., estimator = 'wls'), that will be respected.
    • Some minor adjustments to checking fitted models for specific warnings/errors that should result in better functioning when these do not exist.

CRAN v0.5.5

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 05 Jun 17:13

What's Changed

Full Changelog: v0.5.4...v0.5.5

CRAN v0.5.4

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 22 May 17:50

ShortForm v0.5.4

  • Updates to handle changes to lavaan error/warning messages
  • Fixed internal documentation to match current standards

CRAN v0.5.3

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 14 Dec 14:59

ShortForm v0.5.3

  • Updates to antcolony.lavaan default lavaan.model.specs
  • Spelling fixes
  • Roxygen manual fixes

v0.5.2

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 08 Mar 16:48

CRAN-approved

What's Changed

Full Changelog: v0.4.6...v0.5.2

v0.4.6

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 13 Mar 16:21

ShortForm 0.4.6

Updated argument check

  • The argument check for fit statistics (ACO, SA) were updated to
    include fit statistics related to the “WLSMV” estimator (i.e, for
    ordered data).

Bugfixes

  • A bug in the SA function that sometimes resulted in improper items
    has been fixed.

  • A bug in the Tabu short form function that caused the Tabu list to
    be erased if the best criterion value was reached has been fixed.

v0.4.5

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 25 Feb 15:02
a31b63e

Package v0.4.5

Extra argument checks

  • There are now additional checks for the function arguments related to the fit statistics (ACO, SA) and the fit statistic tests (ACO) with informative warnings for when the function arguments are not valid.

v0.4.4

Choose a tag to compare

@AnthonyRaborn AnthonyRaborn released this 10 Oct 14:35
c9f38db

v0.4.4

  • Bugfixes:
    • I have fixed a bug that caused the lavaan-based functions to improperly use the initial model syntax instead of building the intermediate model syntax when the initial model syntax was specified in an untested manner.
  • Groundwork:
    • Looking to add a reset function to Tabu shortform for when a "dead end" has been found (no valid models exist from the current model). This is implemented here in a very rough, blunt manner. Soon, I want to refactor all the code, especially this implementation, to be more clear and easier to maintain.