Calculate the dispersion parameter for a quasi-poisson model

Converted from STATA from armstrong 2014 supplement with LLM guidance. Minimal error checking of these inputs at present.

Usage

calc_dispersion(y, X, beta, stratum_vector)

Arguments

y

a vector of outcomes

X

a matrix of predictors, typically the crossbasis output

beta

a vector of coefficients

stratum_vector

a vector describing the stratum

Value

a numeric scalar, the estimated dispersion parameter

Examples

dispersion <- calc_dispersion(y = boston_deaths_tbl$daily_deaths,
              X = cb,
              stratum_vector = boston_deaths_tbl$strata,
              beta = coef(m_sub))
#> Error: object 'cb' not found
dispersion
#> Error: object 'dispersion' not found