AccelStab 2.3.1

This is a minor release that changes the following...

step1_down() and step1_sample_mvt()

• Replaced sigma with chisq for multi-temperature models
• The residual error standard deviation (sigma) is now sampled from its own distribution rather than being treated as fixed.

step1_sample_mvt()

• Updated to align with step1_down():
• Added ... (ellipsis) to support optional arguments passed to minpack.lm::nlsLM() during model fitting.

excursion()

• Updated procedure for excursion() to draw from multi-t dist like used in step1_down() and step1_sample_mvt().

AccelStab 2.3.0

This is a minor release that changes the following...

step1_down() and step1_sample_mvt()

• Replaced sigma with chisq for multi-temperature models
• The residual error standard deviation (sigma) is now sampled from its own distribution rather than being treated as fixed.

step1_sample_mvt()

• Updated to align with step1_down():
• Added ... (ellipsis) to support optional arguments passed to minpack.lm::nlsLM() during model fitting.

AccelStab 2.2.0

This is a minor release which adds extra functionality to existing functions plus
a vignette has been added that gives more background to the package.

Enhancements

Vignette

• Added a vignette that gives more information on the background theory of the package.

step1_plot_diagnostic()

• Includes validation data in observed vs predicted and residuals vs predicted scatter plots, if provided.
• Adds a new plot: residuals vs time, color-coded by temperature, and includes validation data points when present.
• Vignette updated to reflect that the function now generates five plots (previously four).

Plotting Functions (step1_plot_CI(), step1_plot_desc(), step1_plot_PI(), step1_plot_pred(), step1_plot_T())

• xlim and ylim now truncate the visible plot area only, without removing data used in the plot.

Bug Fixes

step1_plot_diagnostic()

• Fixed QQ plot to show quantiles for all data, rather than temperature-specific quantiles.

step1_down() and step1_down_basic()

• Added ... to allow passing optional arguments to minpack.lm.
• When reparam = TRUE, now prints a message: default lower bounds are 0, but custom bounds may be specified.
• Fixed issue with Kelvin temperature inputs.

Documentation

• Updated function descriptions across all relevant functions for clarity and completeness.

This is the second version of the AccelStab kinetic modelling package, some extra features are added alongside bug fixes.

New features

• step1_down_rmse() added which allows the user to calculate the root mean squared error for their data and chosen parameters
• step1_down_diagnostic() added which allows the user to plot residual diagnostic plots after fitting the model
• step1_sample_mvt() added which allows the user to draw a chosen number of sample parameters from the multivariate t distribution for their own analyses
• step1_down() now accepts an extra argument validation which sidelines some of the data allowing the user to save it for testing purposes
• step1_plot_desc() now accepts an extra argument validation which sidelines some of the data allowing the user to save it for testing purposes

Bug fixes

• When selecting a temperature outside the data set for step1_plot_T() the colours of the prediction line and the ribbon are now consistent
• When using the argument temp_pred_C within step1_down() no longer are predictions duplicated if the temperature is already in the data
• Issue when no time = 0 rows present in the data and no parms provided to step1_down() now resolved

This is the first release of the Accelerated Stability Kinetic Modelling package! If you encounter any bugs or you would like to see a new feature implemented, then please let us know.