Estimating Skellam distribution and regression parameters in Stata

Vincenzo Verardi, Catherine Vermandele

Research output: Contribution to journalArticlepeer-review

Abstract

The Skellam distribution is a discrete probability distribution related to the difference between two independent Poisson-distributed random variables. It has been used in a variety of contexts, including sports or supply and demand imbalances in shared transportation. Stata does not support the Skellam distribution or Skellam regression. We present a command, skellamreg, to estimate the parameters of a Skellam distribution and Skellam regression model using Mata’s optimize function.

Original languageEnglish
Pages (from-to)287-300
Number of pages14
JournalStata Journal
Volume24
Issue number2
DOIs
Publication statusPublished - Jun 2024

Funding

Vincenzo Verardi gratefully acknowledges the financial support of the Fonds National de la Recherche Scientifique (FNRS).

Funders
Fonds de la Recherche Scientifique F.R.S.-FNRS

    Keywords

    • Skellam distribution
    • Skellam regression
    • maximum likelihood
    • modified Bessel function of the first kind
    • optimize
    • skellamreg
    • skellamreg postestimation
    • st0748

    Fingerprint

    Dive into the research topics of 'Estimating Skellam distribution and regression parameters in Stata'. Together they form a unique fingerprint.

    Cite this