Error, object not declared (allNus?)

[Steve Foley]

Hi all,

I'm using BEAST v1.10.1, and have received the following error:

Creating site rate model.

Creating site rate model.

Creating site rate model.

Creating site rate model.

Creating site rate model.

Creating site rate model.

Creating site rate model.
Sep 04, 2018 1:47:38 PM dr.app.beast.BeastMain <init>
SEVERE: Parsing error - poorly formed BEAST file, NEXTEST.xml:
Object with idref=B62.nu has not been previously declared.

Error thrown at: dr.xml.XMLParser.convert(Unknown Source)

My data is partitioned based on the substitution model I want to apply, so B62 is my Blosum 62 partition. The .nu suggests this is something to do with the allNus parameter, but I'm not sure what to do about it. The only mention of B62.nu in my XML is as follows:

    <!-- Collecting together relative rates for partitions                       -->
    <compoundParameter id="allNus">
        <parameter idref="B62.nu"/>
        <parameter idref="CPR.nu"/>
        <parameter idref="DHF.nu"/>
        <parameter idref="FLU.nu"/>
        <parameter idref="JTT.nu"/>
        <parameter idref="LGM.nu"/>
        <parameter idref="WAG.nu"/>
    </compoundParameter>

Should it be declared somewhere earlier in my XML? Any suggestions would be massively appreciated. Thank you!

[Andrew Rambaut]

Dear Steve,

We will look into this error. In the meantime, perhaps try switching on 'Use classic priors/operators' in the Priors panel.

Andrew

 

--
You received this message because you are subscribed to the Google Groups "beast-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beast-users...@googlegroups.com.
To post to this group, send email to beast...@googlegroups.com.
Visit this group at https://groups.google.com/group/beast-users.
For more options, visit https://groups.google.com/d/optout.

[Rachel]

I'm getting this same error

Rachel

[Rachel]

and I tried your suggestion of using classic prior/operators, but it generated an error 'illegal argument exception: operator with name, UniformPInv, is unknown, and the resulting .xml file is truncated at this section of code.

[Guy Baele]

Can you please e-mail me the full error message that appears on your screen, as well as the XML file you have created, so that we can look into this?
Are you using the latest BEAST version (1.10.4)?

Best regard,
Guy

Op woensdag 2 januari 2019 20:49:34 UTC+1 schreef Rachel:

[Rachel]

of course!

I am using version 1.10.2, as that is the current highest version running on CIPRES.

I am attaching the xml file that triggers the 'not previously declared' error in BEAST (conc_PRV_unlinkedSUBST.xml), as well as the output screen from BEAST when I try to run it (conc_unlinked_error)

[Andrew Rambaut]

Can you try generating the XML in BEAUti v1.10.4 (it should then run on BEAST v1.10.2).

Andrew

<conc_PRV_unlinkedSUBST.xml><conc_unlinked_error>

[Rachel Breyta]

of course! 

I re-created the same analyses in v1.10.4- see xml attached. it did not run in v1.10.2, and when I ran it in v1.10.4, I got the same not previously declared error (beast output also attached).

Rachel B. Breyta, Ph.D.

Sch of Aquatic and Fisheries Sciences, Univ of WA
Microbiology, Oregon State University
Western Fisheries Research Center, USGS
office (206) 526-2277
rbj...@uw.edu, rbr...@usgs.gov, bre...@oregonstate.edu

[Steve Foley]

For context - I still haven't found a way to fix this, so I'm still in the same boat as Rachel.

I've updated to BEAUTi v1.10.4, generated the xml from scratch, and BEAST v1.10.4 terminates with "Object with idref=B62.nu has not been previously declared". Switching on classic priors / operators in BEAUTi opens a prompt saying "Operator with name, uniformPInv is unknown".

Any further ideas on how to fix this would be massively appreciated. Thanks!

S.

[Andrew Rambaut]

Hi Steve,

Are you using amino acids? There are two issues here - the pInv option doesn't work with amino acids in BEAUti (I am not sure when this problem was introduced). Secondly the new relative rates between partitions (the 'nu' parameters) doesn't work with amino acids either. The second problem is got around by selecting 'Classic priors and operators' but the first is not.

We will fix both these issues shortly.

Andrew

[Steve Foley]

Hi Andrew,

Yes, I am using amino acids. Thank you for your kind explanations! Looking forward to further updates.

S.