Recent changes to 31: dweibull() parametrization confusing in the manualhttp://sourceforge.net/p/mcmc-jags/bugs/31/2014-05-21T08:26:09Z#31 dweibull() parametrization confusing in the manual2014-05-21T08:26:09Z2014-05-21T08:26:09ZMartyn Plummerhttp://sourceforge.net/u/userid-1549763/http://sourceforge.net43b14a71c9747d5079c36240dbd365fbc6f8f3ec<div class="markdown_content"><ul>
<li><strong>status</strong>: open --> closed</li>
</ul></div>#31 dweibull() parametrization confusing in the manual2014-05-21T08:25:34Z2014-05-21T08:25:34ZMartyn Plummerhttp://sourceforge.net/u/userid-1549763/http://sourceforge.netee200d5f36a854ea7e2f86a0882c35d32653cb44<div class="markdown_content"><p>Fair point. I added some precision to the section on distribution aliases, noting that using an alias does not change the parameterization. I also point ou the different parameterization of the Weibull distribtuion in BUGS and R.</p>
<p>The problem of incompatible parameterizations is pervasive. For example the normal distribution - which does not need an alias - is parameterized in terms of the precision, not the variance in BUGS. This problem won't go away until I add named parameters for functions and distributions.</p></div>dweibull() parametrization confusing in the manual2014-05-14T14:27:48Z2014-05-14T14:27:48ZAlexey Stukalovhttp://sourceforge.net/u/userid-1529657/http://sourceforge.netca8cf467b96dbb305868698386e36484c23b3f0f<div class="markdown_content"><p>Weibull distribution in JAGS (dweib() in bugs module) is defined in terms of shape and rate parameters. JAGS manual 3.4.0 states that there is dweibull() alias that is compatible with R. This is confusing, because in R (at least 3.0+) dweibull() is defined in terms of shape and scale parameters, and one would expect that in JAGS dweibull() is shape/scale as well. But this is not the case.</p>
<p>If it's not possible to make dweibull() really use shape/scale parametrization, maybe this alias could be removed, as it would force the user to take a second look at the parameters. At the very least the manual should not state R compatibility.</p></div>dweibull() parametrization confusing in the manual2014-05-14T14:27:48Z2014-05-14T14:27:48ZAlexey Stukalovhttp://sourceforge.net/u/userid-1529657/http://sourceforge.net76a25354867f3a1f0b569082bbbf409e46448ae1<div class="markdown_content"><p>Ticket 31 has been modified: dweibull() parametrization confusing in the manual<br />
Edited By: Martyn Plummer (martyn_plummer)<br />
Status updated: u'open' => u'closed'</p></div>