The PLM Procedure

Scoring Data Sets for Zero-Inflated Models

The PLM procedure can score new observations for zero-inflated models with the SCORE statement. If you specify the ILINK option, the computed statistics are for estimated counts.

In the following formula, bold x is the design row for covariates that correspond to the Poisson or negative binomial component, ModifyingAbove bold-italic beta With caret is the column vector of the fitted regression parameters; bold z is the design row for covariates that correspond to the zero inflation component, ModifyingAbove bold-italic gamma With caret is the column vector of the fitted regression parameters; g and gSuperscript negative 1 are the link and inverse link functions for the Poisson or negative binomial component, g Subscript z and g Subscript z Superscript negative 1 are the link and inverse link functions for the zero inflation component; normal upper Phi is the standard normal cumulative distribution function and alpha is the nominal significance level. Let

StartLayout 1st Row 1st Column mu 2nd Column equals 3rd Column g Superscript negative 1 Baseline left-parenthesis eta right-parenthesis 4th Column equals 5th Column g Superscript negative 1 Baseline left-parenthesis bold x ModifyingAbove bold-italic beta With caret right-parenthesis 2nd Row 1st Column mu Subscript z 2nd Column equals 3rd Column g Subscript z Superscript negative 1 Baseline left-parenthesis eta Subscript z Baseline right-parenthesis 4th Column equals 5th Column g Subscript z Superscript negative 1 Baseline left-parenthesis bold z ModifyingAbove bold-italic gamma With caret right-parenthesis 3rd Row 1st Column v 1 2nd Column equals 3rd Column left-parenthesis StartFraction normal d mu Over normal d eta EndFraction right-parenthesis squared bold x ModifyingAbove bold upper V With caret Subscript bold-italic beta comma bold-italic beta Baseline bold x prime 4th Row 1st Column v 2 2nd Column equals 3rd Column left-parenthesis StartFraction normal d mu Subscript z Baseline Over normal d eta Subscript z Baseline EndFraction right-parenthesis squared bold z ModifyingAbove bold upper V With caret Subscript bold-italic gamma comma bold-italic gamma Baseline bold z prime 5th Row 1st Column v 12 2nd Column equals 3rd Column minus left-parenthesis StartFraction normal d mu Over normal d eta EndFraction right-parenthesis left-parenthesis StartFraction normal d mu Subscript z Baseline Over normal d eta Subscript z Baseline EndFraction right-parenthesis bold x ModifyingAbove bold upper V With caret Subscript bold-italic beta comma bold-italic gamma Baseline bold z prime EndLayout

The formula for statistics in the SCORE statement for zero-inflated models are listed as follows.

StartLayout 1st Row 1st Column normal upper P normal upper Z normal upper E normal upper R normal upper O 2nd Column equals 3rd Column omega 4th Column equals 5th Column mu Subscript z 2nd Row 1st Column normal upper P normal upper R normal upper E normal upper D slash normal upper I normal upper L normal upper I normal upper N normal upper K 2nd Column equals 3rd Column p Subscript c 4th Column equals 5th Column mu left-parenthesis 1 minus omega right-parenthesis 3rd Row 1st Column normal upper S normal upper T normal upper D slash normal upper I normal upper L normal upper I normal upper N normal upper K 2nd Column equals 3rd Column s Subscript c 4th Column equals 5th Column StartRoot p Subscript c Superscript 2 Baseline v 2 plus left-parenthesis 1 minus omega right-parenthesis squared v 1 plus v 1 v 2 plus 2 p Subscript c Baseline left-parenthesis 1 minus omega right-parenthesis v 12 plus v 12 squared EndRoot 4th Row 1st Column normal upper U normal upper C normal upper L normal upper M slash normal upper I normal upper L normal upper I normal upper N normal upper K 2nd Column equals 3rd Column u Subscript c 4th Column equals 5th Column p Subscript c Baseline exp left-parenthesis normal upper Phi Superscript negative 1 Baseline left-parenthesis 1 minus alpha slash 2 right-parenthesis s Subscript c Baseline slash p Subscript c Baseline right-parenthesis 5th Row 1st Column normal upper L normal upper C normal upper L normal upper M slash normal upper I normal upper L normal upper I normal upper N normal upper K 2nd Column equals 3rd Column l Subscript c 4th Column equals 5th Column p Subscript c Baseline slash exp left-parenthesis normal upper Phi Superscript negative 1 Baseline left-parenthesis 1 minus alpha slash 2 right-parenthesis s Subscript c Baseline slash p Subscript c Baseline right-parenthesis 6th Row 1st Column normal upper P normal upper R normal upper E normal upper D 2nd Column equals 3rd Column p Subscript l 4th Column equals 5th Column g left-parenthesis p Subscript c Baseline right-parenthesis 7th Row 1st Column normal upper S normal upper T normal upper D 2nd Column equals 3rd Column s Subscript l 4th Column equals 5th Column g left-parenthesis s Subscript c Baseline right-parenthesis slash left-parenthesis StartFraction normal d mu Over normal d eta EndFraction right-parenthesis 8th Row 1st Column normal upper U normal upper C normal upper L normal upper M 2nd Column equals 3rd Column u Subscript l 4th Column equals 5th Column g left-parenthesis u Subscript c Baseline right-parenthesis 9th Row 1st Column normal upper L normal upper C normal upper L normal upper M 2nd Column equals 3rd Column l Subscript l 4th Column equals 5th Column g left-parenthesis l Subscript c Baseline right-parenthesis EndLayout
Last updated: December 09, 2022