1. package: gamlss

- i) The glim.fit() function within gamlss() has a line added to prevent the iterative weighs wt to go to Inf.
- ii) The tp() function within lms() and quantSheets() has changed name and modified slightly
- iii) The vcoc.gamlss() has the warnings changed and allows if theinverse of the Hessian (R) fails to recalucated using different method.
- iv) The output of the term.plot() function has changed.
- v) IMPORTANT: pb() has a new version which is faster than the old one. The old pb() function is renamed and called pbo()
- vi) The summary.gamlss() functions instead of giving an R warning for additive terms it gives a “Note” in the output.
- vii) Q.stats() has been modified to .n.iter=5 will work with small data

2. package: gamlss.demo

- i) The Locmean(), Locpoly(), WLocmean() and WLocpoly() functions were moved from the package gamlss.util to here since they are used only for the demos.
- ii) The function demo.LocalRegression() is added

3. package: gamlss.dist

- i) The discrete distribution double Poisson DPO() is added.
- ii) a modififiacion on plotZAGA() is added

4. package: gamlss.util

- i) garmaFit() is modified to allow binomial (and binary) fits (Thanks to Matthias Schmid for point out to us).
- ii) scattersmooth() the argument “cols” was added to allow different schemes of colours. This also allow the package not to depend to the package colorspace even though the plot is not looking as good. If you have the colorspace in you version of R you can still use the old scheme, see the example in scattersmooth().
- iii) plotSimpleGamlss() : had a minor ammentment thanks to Michael Guan.

5. gamlss.add

- i) the function plotNN() is added for plotting fitted neural netwarks

Version 4.3-1

1. package: gamlss

- i) the functions centiles(), centiles.fan() and centiles.com() had some of the plotting options changed
- ii) the lms() the function has changed with some extra arguments added, and a with an extra predict.lms() method. Now the lms object (created by lms()) can be used with centiles.pred().
- iii) z.scores() function for lms object is added to simplify getting the z.scores for new observations
- iv) the function quantSheets() is added. It uses the method of Schnabel and Eilers (2013) to create quantiles curves for centile estimation. The created object has print(), fitted(), predict() and resid() methods.
- v) the functions z.scoresQS() and findPower() are also added to assist centile estimation modelling using quantSheets().

2. package: gamlss.dist

- i) the function flexDist() is moved from gamlss.util to gamlss.dist
- ii) a bug is fixed in the qZAGA() and qZAIG() functions

Version 4.3-0

1. packages gamlss

- i) kri is added as another smoother
- ii) rvcov() the argument hessian.fun is added
- iii) fp() now saves the lm fitted models and it can be access using getSmo()
- vi) ra() and rc() are gone to be replaced by re() and interface for calling mle() within gamlss

2. package: gamlss.dist

- i) Family() is amended to save a function rather just a list
- ii) all the FORTRAN source files have now translated to C thanks to Marco Enea
- iii) ST3C a version of ST3 written in C rather in R is added thanks to Alexios Ghalanos

3, package gamlss.add

- i) fk() now does not print the results in each iteration (it was left by mistake in the last version)
- ii) ga() ,tr() and nn() now can take bigger formula.

Hi,

Is GAMLSS framework available in Julia? How do you use it in Julia?

Thanks,

Siva

I am afraid no