predict.dfcop: S3 methods for discrete factor copula model

In tvatter/mdmd: Multivariate Discrete Mixture Distributions

Description

Currently available:

predict

Predict the probabilities of new observations.

fitted

Fitted probabilities.

logLik

Log-likelihood at the MLE.

AIC

Akaike's Information Criterion (see AIC).

BIC

Bayesian Information Criterion (see BIC).

coef

The MLE.

vcov

Variance-covariance matrix at the MLE.

Usage

## S3 method for class 'dfcop'
predict(object, newdata, ...)

## S3 method for class 'dfcop'
fitted(object, ...)

## S3 method for class 'dfcop'
logLik(object, ...)

## S3 method for class 'dfcop'
coef(object, ...)

## S3 method for class 'dfcop'
vcov(object, ...)

Arguments

object	a dfcop object.
newdata	points where the fit shall be evaluated.
...	unused.

Examples

# Sample size, dimension, correlation parameter
n <- 1e3
d <- 1e1
rho <- 0.7

# The model
prob <- runif(d)
dfcop <- dfcop_dist(prob, "gaussian", rho)

# Data
X <- rdfcop(n, dfcop)

# Fit
fit <- dfcop(X)
all.equal(predict(fit, X), fitted(fit))

# Other methods
logLik(fit)
AIC(fit)
BIC(fit)

tvatter/mdmd documentation built on May 13, 2019, 4:11 a.m.