Skewness coefficients

#1
Hi
I'm using Gamlss to formulate prediction equations. The problem I am facing is that when the model is normally distributed I can't obtain the skewness coefficients when I use the summary command. How can I get the L or skewness coeffecients. I need them to calculate z score
Sorry if my questions seems to basic... It's the first time I use r
 

Dason

Ambassador to the humans
#2
I don't understand - if you're fitting with a normal distribution then there isn't any skew being modeled.
 
#3
So there is no skewness coefficients right? Another question please, how to display standard error of the estimate of the model?
 
#4
how to determine standard error of the estimate of this model please

> summary(mm.nofvc)
******************************************************************
Family: c("NO", "Normal")

Call: gamlss(formula = fvc ~ log(height) + pb(log(age)),
sigma.formula = ~pb(log(age)), family = NO(mu.link = "log"),
data = mm)

Fitting method: RS()

------------------------------------------------------------------
Mu link function: log
Mu Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -9.88323 0.64845 -15.24 <2e-16 ***
log(height) 2.33865 0.12496 18.71 <2e-16 ***
pb(log(age)) -0.24521 0.01384 -17.72 <2e-16 ***
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

------------------------------------------------------------------
Sigma link function: log
Sigma Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -0.3282 0.2129 -1.541 0.1235
pb(log(age)) -0.1256 0.0595 -2.110 0.0351 *
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

------------------------------------------------------------------
NOTE: Additive smoothing terms exist in the formulas:
i) Std. Error for smoothers are for the linear effect only.
ii) Std. Error for the linear terms maybe are not accurate.
------------------------------------------------------------------
No. of observations in the fit: 1061
Degrees of Freedom for the fit: 7.569118
Residual Deg. of Freedom: 1053.431
at cycle: 4

Global Deviance: 1366.002
AIC: 1381.141
SBC: 1418.736
******************************************************************
>
 
#5
how to determine standard error of the estimate of this model
Sorry, I always hesitates what this expression (“standard error of the estimate ”) means.

If you are referring to the variance in the residuals (in the mean model), then this is given in the equation for the sigma-model. The sigma model says that the log(sigma) is a constant – the intercept – and also that it is influenced by the log(age) (which seems to have a spine function (the “pb( )”).

Here you also have the global deviance and the AIC.

If you want to estimate skewness, I believe that it is possible to estimate an LMS-model, which stand for lambda, mean, standard deviation. Of course lambda represents the skewness.