BEAST v 1.10.1 error

Francisca

unread,

Jul 31, 2018, 9:43:30 AM7/31/18

to beast-users

Hi everyone,

I am trying to run an analysis on beast 1.10 but I am getting the following error:

Using BEAGLE DataLikelihood Delegate

Using BEAGLE resource 0: CPU

with instance flags: PRECISION_DOUBLE COMPUTATJul 31, 2018 10:32:35 AM dr.app.beast.BeastMain <init>

SEVERE: Parsing error - poorly formed BEAST file, newv.xml:

The '<normalPrior>' element, nested within an element with id, 'prior', is incorrectly constructed.

The following was expected:

Any number of ELEMENT of type Statistic REQUIRED

Error thrown at: dr.xml.AbstractXMLObjectParser.parseXMLObject(Unknown Source)

I can run the same analysis on beast 1.8.4 (same dataset, parameters and priors), but when I try to do so on the new version I get this error message. I redid the xml file using beautiful 1.10, but it still won't work. Any ideas?

Cheers,

Francisca

Guy Baele

unread,

Jul 31, 2018, 1:52:51 PM7/31/18

to beast-users

Hi Francisca,

Can you please e-mail me your XML files, both for BEAST 1.8.4 and for BEAST 1.10 ?

I can then have a look to see what's going wrong.

Best regards,
Guy

Op dinsdag 31 juli 2018 06:43:30 UTC-7 schreef Francisca:

David Černý

unread,

Aug 26, 2018, 1:09:41 PM8/26/18

to beast-users

Hi Guy,

I was wondering if this problem was ever resolved: I'm in the same situation (meaning I first imported an XML created using BEAUTi v1.8.4 into BEAUTi v1.10.1 and then exported again) and I'm getting the same error, except that it concerns <exponentialPrior> instead. I'd be happy to send you the relevant files if that helps.

Guy Baele

unread,

Aug 27, 2018, 5:37:29 AM8/27/18

to beast-users

Hi David,

So this is a problem with loading a previously constructed XML file into BEAUti, which isn't BEAUti's main functionality.
I've been able to reproduce the original problem and it's on our list of things to fix (but hasn't been fixed just yet).

Currently, we suggest a workaround, which is to redefine (i.e. remove and add again) all taxon sets after loading an existing BEAST XML into BEAUti.
This should take care of the problem, which is that a prior is being defined on a taxon set, rather than on its tmrca.

Can you please try this and let me know how it goes?

Best regards,
Guy

Op zondag 26 augustus 2018 19:09:41 UTC+2 schreef David Černý:

David Černý

unread,

Aug 27, 2018, 10:14:07 AM8/27/18

to beast-users

Hi Guy,

Many thanks for your quick and helpful reply; redefining the taxon sets in BEAUti did take care of the issue. It also turned out that doing so wasn't actually necessary; it would have sufficed to manually replace the idref="nameoftaxonset" attribute inside the <statistic> element of each calibration prior with idref="tmrca(nameoftaxonset)" in the re-exported XML, since this was the only change that the redefinition introduced into the file.

Best regards,

--

David Černý

Guy Baele

unread,

Aug 27, 2018, 11:57:15 AM8/27/18

to beast-users

Hi David,

The workaround I proposed was mainly meant at fixing the problem by solely using BEAUti, but if you're familiar with editing BEAST XML files, then that's indeed a bit more efficient.
Thanks for posting the XML fix as well.