diagnose_separation: Detailed separation diagnostic for all categorical outcomes.

In divoRce: Diagnosing Separation Phenomena in Latent Polyhedral Categorical Response Models

Detailed separation diagnostic for all categorical outcomes.

Description

This function checks whether there is (quasi-) complete separation, which type if any, gives the dimension of the recession cone, lists the number of columns in the design matrix that give rise to the separation as well as the columns names and lists the rows in X/S for which we have separation.

Usage

diagnose_separation(object, ...)

## Default S3 method:
diagnose_separation(object, ...)

## S3 method for class 'factor'
diagnose_separation(
  object,
  X,
  model = NULL,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'character'
diagnose_separation(
  object,
  X,
  model = NULL,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'logical'
diagnose_separation(
  object,
  X,
  model = NULL,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'numeric'
diagnose_separation(
  object,
  X,
  model = NULL,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'integer'
diagnose_separation(
  object,
  X,
  model = NULL,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'matrix'
diagnose_separation(
  object,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'formula'
diagnose_separation(
  object,
  data,
  model = NULL,
  rational = FALSE,
  contrasts = NULL,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'osm'
diagnose_separation(
  object,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'clm'
diagnose_separation(
  object,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'polr'
diagnose_separation(
  object,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'multinom'
diagnose_separation(
  object,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'glm'
diagnose_separation(
  object,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'bracl'
diagnose_separation(
  object,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

## S3 method for class 'brmultinom'
diagnose_separation(
  object,
  rational = FALSE,
  backend = c("rcdd", "ROI"),
  solver = NULL,
  ...
)

Arguments

object	an R object. Supported inputs: Pre-fit (vector): a vector of type factor, character, logical, numeric or integer. In this case one also needs to supply the argument X and, optionally but recommended, a model. Pre-fit (matrix): a matrix of structure vectors. Pre-fit (formula): an object of class "formula" (or one that can be coerced to that class): a symbolic description of the model to be fitted. The details of model specification are given under ‘Details’ in glm. In this case one needs to supply data, with an optional model argument as well. Post-fit: a fitted model object of class glm, polr, clm, osm, brmultinom, bracl, brglm or multinom.
...	further arguments passed to the methods or lower-level functions.
X	a design matrix, e.g. generated via a call to model.matrix. X is expected to already contain the desired contrasts for factors (e.g., dummies) and any other expanded columns (e.g., for polynomials).
model	character string specifying the model class to be fitted. One of: "b" – binary "bcl" – baseline-category link "cl" – cumulative link "acl" – adjacent-category link "sl" – sequential link "os" – ordered stereotype model If missing or NULL, defaults to "cl" for ordinal y and "bcl" otherwise.
rational	logical; should rational arithmetic be used?
backend	character string specifying the backend for the linear program. One of "rcdd" (default, and the only option when rational = TRUE) or "ROI".
solver	character string specifying the solver used by the backend. Defaults to "DualSimplex" for backend = "rcdd", and to the first LP solver returned by ROI_applicable_solvers for backend = "ROI".
data	either a standard data frame, list or environment (or object coercible by as.data.frame to a data frame) containing the variables in the model. If not found in data, the variables are taken from environment(formula), typically the environment from which the function is called. Alternatively, data can be a data frame or matrix containing rational numbers as per the definition in rcdd (i.e. columns are characters, entries are either integer numbers or ratios of integer numbers, e.g. "1" or "-234/19008"). This is checked internally; see ‘Details’ for what happens when this structure is discovered.
contrasts	an optional list. See the contrasts.arg of model.matrix.default. Only effective for standard data frames.

Details

diagnose_separation is an S3 generic function.

For developers: If a method should be provided for the generic, it is best to have that method create a matrix of structure vectors S and use the generic with the matrix.

The formula method is for standard data frames and formulas that work the same way as when used with glm. It does not support extended formulas, and may not work for functions that do formula processing differently.

For a data frame or matrix given as rational numbers in the rcdd definition, this is recognized but the formula does not get expanded and is taken literally. Consequently:

• variables in formula must match exactly with the column names in data;

• factors need to be converted to dummies beforehand (which would not be possible in the rational format in any other way anyway).

Value

an object of class 'sepmod' that is a list with the components:

• separation boolean whether there is separation ('TRUE' means separation)

• septype which type of separation (or not). A string of either "Overlap", "Quasi-Complete Separation" or "Complete Separation".

• reccdim dimension of recession cone

• offrows rows associated with separation

• nr.offcols number of columns of the design matrix that have separation

• offcols columns associated with separation

For model = "sl" an object of class 'sepmod_sl'. It is a list or the above lists with the elements corresponding to each category.

Examples

data(qcsepdatm)

## Binary data
data(csepdat1)
outc<-csepdat1$y
desma<-cbind("(Intercept)"=1,csepdat1[,2:ncol(csepdat1)])
c1<- diagnose_separation(outc, X = desma, model = "b")
print(c1)
print(c1, "full")

# Nominal data 
# Baseline-category link
data(qcsepdatm)
y<-qcsepdatm$y
X<-cbind(1,qcsepdatm[,2:ncol(qcsepdatm)])
diagnose_separation(y, X = X, model = "bcl")

# Ordinal data
data(qcsepdato)
yo<-qcsepdato$y
Xo<-qcsepdato[,2:ncol(qcsepdato)]

# Sequential link
diagnose_separation(yo, Xo, model = "sl")

# Ordered stereotype
diagnose_separation(yo, Xo, model = "os")

# Adjacent-category link
diagnose_separation(yo, Xo, model= "acl")

# Cumulative link
diagnose_separation(yo, Xo, model = "cl")

#post fit
if (require('nnet')) {
m1 <- nnet::multinom(y ~ x1 + x2, data = qcsepdatm)
diagnose_separation(m1)
}

divoRce documentation built on Oct. 18, 2026, 5:08 p.m.