rpm
The
R software package
rpm:
that estimates revealed preferences based on observed bipartite matchings as
detailed in
Goyal, Handcock, Jackson, Rendall and Yeung (2023).
Many problems in demography require models for partnership formation that
separate latent preferences for partners from the availability of partners.
This package consider models for matchings within a bipartite population where
individuals have utility for people based on observed and unobserved characteristics.
People can form a partnership or remain unpartnered. The models represent both
the availability of potential partners of different types and preferences of
individuals for such people.
We develop a large-population likelihood framework to estimate
preference parameters based on sample survey data on partnerships and population
composition. The framework was originally due to Dagsvik (2000) and Menzel (2015).
Based on simulation studies conducted in
Goyal, Handcock, Jackson, Rendall and Yeung (2023)
for realistic population sizes, the model recovers preference parameters that are
invariant under different population availabilities. The software uses bootstrap methods to bias
correct parameter estimates for small population sizes and produce confidence intervals
that have the correct coverage.
The CRAN
manual is very detailed.
For more information, click
here.