Changelog
Source:NEWS.md
mmrm 0.3.8
CRAN release: 20240124
New Features

Anova
is implemented formmrm
models and available upon loading thecar
package. It supports type II and III hypothesis testing.  The argument
start
formmrm_control()
is updated to allow better choices of initial values. 
confint
onmmrm
models will give tbased confidence intervals now, instead of the normal approximation.
Bug Fixes
 Previously if the first optimizer failed, the best successful fit among the remaining optimizers was not returned correctly. This is fixed now.
Miscellaneous
 In documentation of
mmrm_control()
, the allowedvcov
definition is corrected to “EmpiricalJackknife” (CR3), and “EmpiricalBiasReduced” (CR2).  Fixed a compiler warning related to missing format specification.
 If an empty contrast matrix is provided to
df_md
, it will return statistics withNA
values.
mmrm 0.3.7
CRAN release: 20231208
New Features
 The argument
method
ofmmrm()
now only specifies the method used for the degrees of freedom adjustment.  Add empirical, empirical Jackknife and empirical biasreduced adjusted coefficients covariance estimates, which can be specified via the new
vcov
argument ofmmrm()
.  Add residual and betweenwithin degrees of freedom methods.
 Add KenwardRoger support for spatial covariance structures.
 Add
model.matrix()
andterms()
methods to assist in postprocessing.  Add
predict()
method to obtain conditional mean estimates and prediction intervals.  Add
simulate()
method to simulate observations from the predictive distribution.  Add
residuals()
method to obtain raw, Pearson or normalized residuals.  Add
tidy()
,glance()
andaugment()
methods to tidy the fit results into summary tables.  Add
tidymodels
framework support via aparsnip
interface.  Add argument
covariance
tommrm()
to allow for easier programmatic access to specifying the model’s covariance structure and to expose covariance customization through thetidymodels
interface.
Bug Fixes
 Previously
mmrm()
follows the global optionna.action
and if it is set other than"na.omit"
an assertion would fail. This is now fixed and henceNA
values are always removed prior to model fitting, independent of the globalna.action
option.  Previously a
model.frame()
call on anmmrm
object with transformed terms, or new data, e.g.model.frame(mmrm(Y ~ log(X) + ar1(VISITID), data = new_data)
, would fail. This is now fixed.  Previously
mmrm()
always required adata
argument. Now fittingmmrm
can also use environment variables instead of requiringdata
argument. (Note thatfit_mmrm
is not affected.)  Previously
emmeans()
failed when using transformed terms or not including the visit variable in the model formula. This is now fixed.  Previously
mmrm()
might provide nonfinite values in the Jacobian calculations, leading to errors in the Satterthwaite degrees of freedom calculations. This will raise an error now and thus alert the user that the model fit was not successful.
Miscellaneous
 Use automatic differentiation to calculate Satterthwaite adjusted degrees of freedom, resulting in 10fold speedup of the Satterthwaite calculations after the initial model fit.
 Add an interactive confirmation step if the number of visit levels is too large for nonspatial covariance structures. Use
options(mmrm.max_visits = )
to specify the maximum number of visits allowed in noninteractive mode.  Removed
free_cores()
in favor ofparallelly::availableCores(omit = 1)
.  The
model.frame()
method has been updated: Thefull
argument is deprecated and theinclude
argument can be used instead; by default all relevant variables are returned. Furthermore, it returns adata.frame
the size of the number of observations utilized in the model for all combinations of theinclude
argument whenna.action= "na.omit"
.  Overall, seven vignettes have been added to the package. All vignettes have a slightly different look now to reduce the size of the overall R package on CRAN.
 The used optimizer is now available via
component(., "optimizer")
instead of previouslyattr(., "optimizer")
.
mmrm 0.2.2
CRAN release: 20221220
New Features
 Add support for KenwardRoger adjusted coefficients covariance matrix and degrees of freedom in
mmrm
function call with argumentmethod
. Options are “KenwardRoger”, “KenwardRogerLinear” and “Satterthwaite” (which is still the default). Subsequent methods calls will respect this initial choice, e.g.vcov(fit)
will return the adjusted coefficients covariance matrix if a KenwardRoger method has been used.  Update the
mmrm
arguments to allow users more finegrained control, e.g.mmrm(..., start = start, optimizer = c("BFGS", "nlminb"))
to set the starting values for the variance estimates and to choose the available optimizers. These arguments will be passed to the new functionmmrm_control
.  Add new argument
drop_visit_levels
to allow users to keep all levels in visits, even when they are not observed in the data. Dropping unobserved levels was done silently previously, and now a message will be given. See?mmrm_control
for more details.
Bug Fixes
 Previously duplicate time points could be present for a single subject, and this could lead to segmentation faults if more than the total number of unique time points were available for any subject. Now it is checked that there are no duplicate time points per subject, and this is explained also in the function documentation and the introduction vignette.
 Previously in
mmrm
calls, theweights
object in the environment where the formula is defined was replaced by theweights
used internally. Now this behavior is removed and your variableweights
e.g. in the global environment will no longer be replaced.
mmrm 0.1.5
CRAN release: 20221018
 First CRAN version of the package.
 The package fits mixed models for repeated measures (MMRM) based on the marginal linear model without random effects.
 The motivation for this package is to have a fast, reliable (in terms of convergence behavior) and feature complete implementation of MMRM in R.
New Features
 Currently 10 covariance structures are supported (unstructured; as well as homogeneous and heterogeneous versions of Toeplitz, autoregressive order one, antedependence, compound symmetry; and spatial exponential).
 Fast C++ implementation of Maximum Likelihood (ML) and Restricted Maximum Likelihood (REML) estimation.
 Currently Satterthwaite adjusted degrees of freedom calculation is supported.
 Interface to the
emmeans
package for computing estimated marginal means (also called leastsquare means) for the coefficients.  Multiple optimizers are run to reach convergence in as many cases as possible.
 Flexible formula based model specification and support for standard S3 methods such as
summary
,logLik
, etc.