create_bseq: Creates a list of data for a sequence of binary models from a...

In divoRce: Diagnosing Separation Phenomena in Categorical Response Models

Creates a list of data for a sequence of binary models from a sequential (continuation ratio) model. It splits the data according to the forward sequential mechanism.

Description

Creates a list of data for a sequence of binary models from a sequential (continuation ratio) model. It splits the data according to the forward sequential mechanism.

Usage

create_bseq(y, X)

Arguments

y	the ordinal outcome variable with m categories. Works best if it is a factor but can also be numeric, boolean or character. In the later case we interpret the ordering as alphanumerically increasing (just like as.ordered is doing).
X	a design matrix, e.g. generated via a call to 'model.matrix'. This means we expect that X already contains the desired contrasts for factors (e.g., dummies) and any other expanded columns (e.g., for polynomials).

Details

The sequential mechanism is so that or category k (running from 1 to m-1) all i for which y_i>=k are selected. All y_i>k are given a value of 0 and y_i=k a value of 1.

Value

a list of length m-1 that separates the input into y and X tupels according to the sequential model mechanism.

divoRce documentation built on April 28, 2026, 3:01 a.m.