Entirely rebuilt using the S7 object-oriented
system (successor to S3/S4) for improved performance and
maintainability.
Core Class Refactoring: The Optimization, Evaluation, and
plotEvaluation classes have been redesigned as fully compliant S7
objects, resulting in a more modular and stable codebase.
Strategic S7 Refactoring: The Optimization, Evaluation, and
plotEvaluation classes have been prioritized for full S7 compliance,
serving as the cornerstone of the package’s new modular architecture to
ensure maximum stability and rigorous property validation.
Breaking Changes
Significant updates required for user scripts from version 6.1.
Changed outcomes slot to outputs in
ModelError and Evaluation classes.
Modified FIM type specification from fim to
fimType.
Updated ODE initial conditions: use "dose_C1" instead
of "dose".
Administration constraints for doses now require a
list() instead of a vector().
Sensitivity indices plot method renamed from
plotSensitivityIndice to
plotSensitivityIndices.
SE and RSE extraction simplified (direct access without
[[designName]] indexing).
New plot() methods for comprehensive design assessment, including
model output evaluation, parameter sensitivity indices, SE and RSE.
All user-facing functions and methods in PFIM now include executable
examples. All exported functions are now documented with @examples sections that
illustrate primary use cases in a clear and reproducible manner
Minor changes
Naming Consistency: Renamed proportionsOfsubjects to
proportionsOfSubjects to strictly adhere to the UpperCamelCase
convention used throughout the PFIM package.
New Features
Flexible PK/PD Modeling: Full support for multiple
PK responses with PD components (ODE-based).
Library of Models: Introduction of
modelFromLibrary for quick access to predefined PK/PD
models.
Bayesian FIM: Enhanced estimation for single
sampling time scenarios.
Multi-dose Bolus: New ODE-based modeling for
complex dosing regimens.
Documentation: Completely redesigned Vignettes 1
and 2 with step-by-step guides for evaluation and optimization.
Deprecated and Removed
Version Deprecation: All versions of PFIM prior to 7.0.3 are now
officially deprecated. Users are encouraged to migrate to the current
release for improved stability.
Report Streamlining: The correlation matrix has been removed from
the standard report to simplify output, though diagnostic data remains
accessible via specialized functions.