dm_poibin: Create a data model based on a Poisson-binomial mixture
In ONSdigital/Bayesian-demographic-accounts: Bayesian Demographic Accounts

Description Usage Arguments Value Examples

Create a data model where the distribution of the reported value is a Poisson-binomial mixture. If Y ~ binom(n, p) and Z ~ pois(n(1-p)) then X = Y + Z is a Poisson-binomial mixture.

1	dm_poibin(data, prob, nm_series, nm_data = NULL)

`data`	A data frame, described in `data-arg`.
`prob`	A number between 0 and 1.
`nm_series`	The name of the demographic series that `data` describes.
`nm_data`	The name of the dataset. If no value supplied, then `nm_data` is assumed to equal the name of the object supplied as the `data` argument.

An object of class "dm_poibin".

reg_immig <- account::gl_reg_immig
reg_immig_dm <- dm_poibin(data = reg_immig,
                          prob = 0.95,
                          nm_series = "immigration")
reg_immig_dm