-
Fixed the output of
predFit()
in situations whenever standard errors and confidence/predictions bands are both requested. -
New data sets
bladder
(a repeated measures data set) andwhisky
. -
Fixed typos in documentation throughout the package.
-
Updated URLs throughout the package.
-
Better y-axis limits when x-axis is on log scale (e.g., using log = "x"). Fixes issue #27.
-
Better default y-axis label when using
plotFit()
on a model with transformed response. For example, callingplotFit(fit)
wherefit <- lm(sqrt(dist) ~ speed, data = cars)
will have a default y-axis label of"sqrt(dist)"
. -
plotFit()
has been completely re-written using much less code. -
predFit()
(and henceplotFit()
) now works for"nls"
objects fit using the Golub-Pereyra algorithm (i.e.,algorithm = "plinear"
); however, confidence/prediction bands are still not available. -
New introductory vignette.
-
Multiple predictor variables are allowed for
"lm"
and"glm"
objects. -
All non-base package functions are now imported.
-
The generic function
predFit()
is now exported. This function is used by investr to obtain predictions, and hence, inverse predictions. For example,predFit()
can be used to obtain prediction intervals for nonlinear least-squares fits (i.e., models of class"nls"
). -
Improved tests and test coverage.
-
plotFit()
gained methods for"rlm"
and"lqs"
objects from package MASS.
-
invest()
now accepts objects of class"glm"
(experimental). -
Functions
calibrate()
andinvest()
now return an object of class"invest"
.
-
Cleaned up documentation.
-
Added
AnyNA()
function for those using older versions of R.
-
Cleaned up examples.
-
Added
bootstrap
option toinvest()
.
- Changed tests to satisfy CRAN check.
-
Updated citation file.
-
Minor code changes.
-
plotFit()
should now plot models with transformed responses correctly. -
Fixed error causing
invest()
to fail because of a missing data argument. -
Added more tests.
-
invest()
now accepts objects of class"lme"
(experimental). -
A few minor bug fixes and code improvements.
-
Added more tests.
-
A few minor bug fixes.
-
Slightly better documentation.
- Initial release.