Some EM-type algorithms to estimate parameters for the well-known Heckman selection model are provided in the package. Such algorithms are as follow: ECM(Expectation/Conditional Maximization), ECM(NR)(the Newton-Raphson method is adapted to the ECM) and ECME(Expectation/Conditional Maximization Either). Since the algorithms are based on the EM algorithm, they also have EM’s main advantages, namely, stability and ease of implementation. Further details and explanations of the algorithms can be found in Zhao et al. (2020) <doi:10.1016/j.csda.2020.106930>.
| Version: | 1.0.1 |
| Depends: | R (≥ 2.10) |
| Imports: | sampleSelection, mvtnorm |
| Published: | 2020-05-31 |
| Author: | Kexuan Yang, Sang Kyu Lee, Zhao Jun, and Hyoung-Moon Kim |
| Maintainer: | Sang Kyu Lee <leesa111 at msu.edu> |
| License: | GPL-2 |
| NeedsCompilation: | no |
| CRAN checks: | EMSS results |
| Reference manual: | EMSS.pdf |
| Package source: | EMSS_1.0.1.tar.gz |
| Windows binaries: | r-devel: EMSS_1.0.1.zip, r-release: EMSS_1.0.1.zip, r-oldrel: EMSS_1.0.1.zip |
| macOS binaries: | r-release: EMSS_1.0.1.tgz, r-oldrel: EMSS_1.0.1.tgz |
| Old sources: | EMSS archive |
Please use the canonical form https://CRAN.R-project.org/package=EMSS to link to this page.