Crandore Hub

DIRMR

Distributed Imputation for Random Effects Models with Missing Responses

By adding over-relaxation factor to PXEM (Parameter Expanded Expectation Maximization) method, the MOPXEM (Monotonically Overrelaxed Parameter Expanded Expectation Maximization) method is obtained. Compare it with the existing EM (Expectation-Maximization)-like methods. Then, distribute and process five methods and compare them, achieving good performance in convergence speed and result quality.The philosophy of the package is described in Guo G. (2022) <doi:10.1007/s00180-022-01270-z>.

Versions across snapshots

VersionRepositoryFileSize
0.5.0 2026-04-09 windows/windows R-4.5 DIRMR_0.5.0.zip 122.5 KiB

Dependencies (latest)

Imports