NEWS.md

In jsnwyse/SimVitD: Simulation Tools for Planning Vitamin D Studies

SimVitD 2.0

• New C implementation of simulation included (faster)
• Bootstrap BCa method from wBoot translated to C code to be accessed directly during simulation
• Added new internal function power.calc.x() to access the C implementation

SimVitD 1.0.2

• Added functions from wBoot by Neil A Weiss into package due to archival
• Due to wBoot incorporation dependency changed to simpleboot
• Included CLT approximation to be used when larger samples
• Determination of CLT/bootstrap is carried out automatically- clt argument to power.calc can be used to set
• Change default number of bootstrap re-samples to 9999

SimVitD 1.0.1

Minor changes

• Added extra argument start to exposure.levels() to pass the time unit in years of study start
• Corrected bug in time offset of exposure levels when start > 0
• Changed default plotting label for 25OHD from '25 Hydroxy Vitamin D' to '25-hydroxyvitamin D'
• Added new function 'vitd.curve.2.function' which will convert the output of vitd.curve into a list of callable functions (not available at user level)

SimVitD 1.0.0

Major changes

• Generative model of vitamin D status curves from a squared sine to cosine with threshold.
• Rather than maximum and minimum levels of vitamin D a more flexible mean level is now used.
• Changes to arguments passed to vitd.curves().
• Cross-over designs have been disabled in current version.
• Parallel computation is now available for Monte Carlo error quantification in pow.calc().
• Addition of internal function power.calc.0().
• prop.test() and wilcox.test() used in version 0.1.2 replaced by non-parametric bootstrap from wBoot package.
• Dependency on poisson package removed as it was possible to vectorize our own sampling.
• Updated print and summary function for exposure.levels and infection.count objects.
• Changes to package vignette to document user facing changes.

jsnwyse/SimVitD documentation built on April 21, 2022, 10:07 a.m.