Skip to content

LALPulsar LineRobustStatsTest check fails on conda

Description of problem

The LineRobustStatsTest check fails when built during conda-forge CI against the latest lal builds. I cannot reproduce this locally, which confuses me no end. It looks like some sort of overflow may be happening, based on the numbers in the test log.

https://dev.azure.com/conda-forge/feedstock-builds/_build/results?buildId=136809&view=logs&j=1bf226d3-0e2f-52d8-fa93-7d9e633347b3&t=ec8d466d-e3b4-5115-4f06-222398f91e6c

If I look at the log for a passing test run on Hawk (Cardiff):

Starting coherent test...
Computing BSGL for TwoF_multi=7.000000, TwoFX=(4.000000,12.000000), priors cohF*0=-340282346638528859811704183484516925440.000000 and oLtLGX=(0.500000,0.800000)...
Externally recomputed     with  allterms: log10LRS=-0.632042
Externally recomputed     with !allterms: log10LRS=-0.627098
XLALComputeBSGL()         with  allterms: log10LRS=-0.632042 (rel. dev.: -0.000000) ==> OK!
XLALComputeBSGL()         with !allterms: log10LRS=-0.627098 (rel. dev.: -0.000000) ==> OK!
Computing BSGL for TwoF_multi=7.000000, TwoFX=(4.000000,12.000000), priors cohF*0=10.000000 and oLtLGX=NULL...
Externally recomputed     with  allterms: log10LRS=-2.525915
Externally recomputed     with !allterms: log10LRS=-2.521884
XLALComputeBSGL()         with  allterms: log10LRS=-2.525915 (rel. dev.: -0.000000) ==> OK!
XLALComputeBSGL()         with !allterms: log10LRS=-2.521884 (rel. dev.: -0.000000) ==> OK!
Starting semi-coherent test...
Computing semi-coherent BSGL for sum_TwoF_multi=16.000000, sum_TwoFX=(37.656853,11.313708), priors cohF*0=10.000000 and oLtLGX=NULL...
Externally recomputed     with  allterms: log10LRS=-9.253451
Externally recomputed     with !allterms: log10LRS=-9.253448
XLALComputeBSGL()         with  allterms: log10LRS=-9.253451 (rel. dev.: -0.000000) ==> OK!
XLALComputeBSGL()         with !allterms: log10LRS=-9.253448 (rel. dev.: -0.000000) ==> OK!

However, on conda-forge Azure CI:

Starting coherent test...
Computing BSGL for TwoF_multi=7.000000, TwoFX=(4.000000,12.000000), priors cohF*0=-340282346638528859811704183484516925440.000000 and oLtLGX=(0.500000,0.800000)...
Externally recomputed     with  allterms: log10LRS=-0.632042
Externally recomputed     with !allterms: log10LRS=-0.627098
XLALComputeBSGL()         with  allterms: log10LRS=-0.632042 (rel. dev.: -0.000000) ==> OK!
XLALComputeBSGL()         with !allterms: log10LRS=-0.627098 (rel. dev.: -0.000000) ==> OK!
Computing BSGL for TwoF_multi=7.000000, TwoFX=(4.000000,12.000000), priors cohF*0=10.000000 and oLtLGX=NULL...
Externally recomputed     with  allterms: log10LRS=-2.525915
Externally recomputed     with !allterms: log10LRS=-2.521884
XLALComputeBSGL()         with  allterms: log10LRS=-2.525915 (rel. dev.: -0.000000) ==> OK!
XLALComputeBSGL()         with !allterms: log10LRS=-2.521884 (rel. dev.: -0.000000) ==> OK!
Starting semi-coherent test...
Computing semi-coherent BSGL for sum_TwoF_multi=16.000000, sum_TwoFX=(669322611770603337440946540249088.000000,70370243876339914375168.000000), priors cohF*0=10.000000 and oLtLGX=NULL...
Externally recomputed     with  allterms: log10LRS=-inf
Externally recomputed     with !allterms: log10LRS=-inf
XLALComputeBSGL()         with  allterms: log10LRS=-145341559238119685803706385694720.000000 (rel. dev.: -nan) ==> BAD!

Note the sum_TwoFX numbers, which seem irrationally large in this context.

Expected behaviour

The test passes?

Steps to reproduce

Not sure.

Context/environment

This job was run in conda-forge Azure CI here

Conda info 
$ conda info
     active environment : base
    active env location : /opt/conda
            shell level : 1
       user config file : /home/conda/.condarc
 populated config files : /opt/conda/.condarc
                          /home/conda/.condarc
          conda version : 4.8.3
    conda-build version : 3.19.1
         python version : 3.7.6.final.0
       virtual packages : __glibc=2.12
       base environment : /opt/conda  (writable)
           channel URLs : https://conda.anaconda.org/conda-forge/linux-64
                          https://conda.anaconda.org/conda-forge/noarch
                          https://repo.anaconda.com/pkgs/main/linux-64
                          https://repo.anaconda.com/pkgs/main/noarch
                          https://repo.anaconda.com/pkgs/r/linux-64
                          https://repo.anaconda.com/pkgs/r/noarch
          package cache : /opt/conda/pkgs
                          /home/conda/.conda/pkgs
       envs directories : /opt/conda/envs
                          /home/conda/.conda/envs
               platform : linux-64
             user-agent : conda/4.8.3 requests/2.23.0 CPython/3.7.6 Linux/4.15.0-1071-azure centos/6.10 glibc/2.12
                UID:GID : 1001:1001
             netrc file : None
           offline mode : False

Suggested solutions