NEWS.md

In IrregLong: Analysis of Longitudinal Data with Irregular Observation Times

IrregLong 0.3.4

Minor changes

• Added error message to extent.of.irregularity for when formula and/or ncutpts are not specified in the absence of scheduled visit times
• made use of data.table package when lagging variables and adding censored rows

Bug fixes

• extent.of.irregularity previously only allowed the time variable to be called "time"; this has now been fixed

IrregLong 0.3.3

Minor changes

• due to archival of the frailtypack package, frailty models are now estimated using coxph rather than frailtyPenal

IrregLong 0.3.2

Bug fixes

• iiw.weights failed to consider the interval between last observed visit and end of follow-up time when maxfu was supplied as a single number. This has been fixed.

IrregLong 0.3.1

Bug fixes

• Fixed a coding error that led extent.of.irregularity to fail when scheduledtimes was provided.

IrregLong 0.3.0

Major changes

• Added function (extent.of.irregularity) to provide graphical and numeric summaries of the extent of irregularity.

Minor changes

• Added an option in the Liang function to provide as an argument a function computing the value of the covariates X for any subject at any given time t, not just the observation times.

IrregLong 0.2.0

Major changes

• When using lagged variable, there is an option to include a value for the first lagged observation within each subject.
• A function (abacus.plot) for producing an abacus plot has been added.
• The Liang model now allows for time-invariant covariates to be included in the visit process model.

Bug fixes

• The documentation (examples and vignette) has been updated to use dosing by weight in the measurement frequency model.
• tinvarcols can be omitted from addcensoredrows, iiwgee, and iiw.weights if maxfu=NULL.
• lagfn and addcensoredrows no longer require subject identifiers to be numeric.

IrregLong documentation built on March 18, 2022, 5:45 p.m.