Extract Model Coefficients

Extracts the estimated coefficients from VLM objects such as VGLMs.

coefvlm(object, matrix.out = FALSE, label = TRUE, colon = FALSE, ...)

Arguments

object

An object for which the extraction of coefficients is meaningful. This will usually be a vglm object.

matrix.out

Logical. If TRUE then a matrix is returned. The explanatory variables are the rows. The linear/additive predictors are the columns. The constraint matrices are used to compute this matrix.

label

Logical. If FALSE then the names of the vector of coefficients are set to NULL.

colon

Logical. Explanatory variables which appear in more than one linear/additive predictor are labelled with a colon, e.g., age:1, age:2. However, if it only appears in one linear/additive predictor then the :1 is omitted by default. Then setting colon = TRUE will add the :1.

...

Currently unused.

Details

This function works in a similar way to applying coef() to a lm or glm object. However, for VGLMs, there are more options available.

Value

A vector usually. A matrix if matrix.out = TRUE.

References

Yee, T. W. and Hastie, T. J. (2003). Reduced-rank vector generalized linear models. Statistical Modelling, 3, 15–41.

Author

Thomas W. Yee

See also

vglm, coefvgam, coef.

Examples

zdata <- data.frame(x2 = runif(nn <- 200))
zdata <- transform(zdata, pstr0  = logitlink(-0.5 + 1*x2, inverse = TRUE),
                          lambda =   loglink( 0.5 + 2*x2, inverse = TRUE))
zdata <- transform(zdata, y2 = rzipois(nn, lambda, pstr0 = pstr0))

fit2 <- vglm(y2 ~ x2, zipoisson(zero = 1), data = zdata, trace = TRUE)
#> Iteration 1: loglikelihood = -326.12618
#> Iteration 2: loglikelihood = -324.79101
#> Iteration 3: loglikelihood = -324.70759
#> Iteration 4: loglikelihood = -324.69898
#> Iteration 5: loglikelihood = -324.69804
#> Iteration 6: loglikelihood = -324.69793
#> Iteration 7: loglikelihood = -324.69791
#> Iteration 8: loglikelihood = -324.69791
coef(fit2, matrix = TRUE)  # Always a good idea
#>             logitlink(pstr0) loglink(lambda)
#> (Intercept)      -0.09140163       0.3102218
#> x2                0.00000000       2.1841194
coef(fit2)
#> (Intercept):1 (Intercept):2            x2 
#>   -0.09140163    0.31022181    2.18411941 
coef(fit2, colon = TRUE)
#> (Intercept):1 (Intercept):2          x2:1 
#>   -0.09140163    0.31022181    2.18411941