NEWS

---

MECfda 0.2.0

Maintainer: Heyang Ji

Previous Version: 0.1.0

New Features

• Added classes and functions required for basis expansion using Functional Principal Component (FPC) basis, including:
  • s4 classes: numeric_basis, numericBasis_series.
  • s4 class methods: plot.numeric_series.
  • Functions: FPC_basis_expansion(), numeric_basis_expansion(), numericBasisSeries2fun().

Enhancements & Compatibility Improvements

• Enhanced existing functions to support the newly introduced FPC basis expansion:
  • Added an option to use the FPC basis in scalar-on-function regression functions, fcRegression and fcQR.
  • Updated S3 methods predict.fcRegression and predict.fcQR to support FPC basis.
  • Extended the generic function basis2fun with new methods.

New Utility Functions

• Introduced two functions for generating synthetic test datasets:
  • MECfda_simDataGen_fcReg
  • MECfda_simDataGen_ME

Bug Fixes

• Fixed an issue in bspline_basis_expansion and fourier_basis_expansion where the numerical integration step did not properly handle non-default integration intervals.

Other Improvements

• Added the function MEM_X_hat(), which returns $\hat X(t)$ used in ME.fcRegression_MEM.
• Removed the package dependency on stringr.
• Updated and refined the help documentation and vignette content.
• Corrected author roles in the author list.