[Users] problems with AHFinderDirect

[Ian Hinder]

On 14 Jun 2012, at 16:05, Erik Schnetter wrote:

> On Thu, Jun 14, 2012 at 8:40 AM, Ian Hinder <ian.hinder at aei.mpg.de> wrote:
>> 
>> On 14 Jun 2012, at 13:53, Eloisa Bentivegna wrote:
>> 
>>> On Jun 14, 2012, at 12:01 PM, Vassilios Mewes wrote:
>>> 
>>>> Hello all,
>>>> 
>>>> when using the AHFinderDirect thorn in a simulation of a relativistic wind past a TOV neutron star, i get the following error message:
>>>> 
>>>> INFO (AHFinderDirect): proc 0: searching for horizon 1/1
>>>>  [1mWARNING level 0 in thorn AHFinderDirect processor 0 host permeability.daa.uv.es
>>>>  (line 78 of /Users/vass/programs/Cactus/arrangements/EinsteinAnalysis/AHFinderDirect/src/jtutil/error_exit.cc):
>>>>  -> [0m ***** row_sparse_Jacobian__UMFPACK::solve_linear_system():
>>>>        error return status=1 from umfpack_numeric() routine
>>>> 
>>>> is that an indication that umfpack is not installed correctly (i was never prompted to install it manually and the compilation of the Einstein toolkit [Maxwell release] went on smoothly without any errors), or rather that something is wrong trying to find the horizon? (there should be an horizon, because the lapse goes to 0 fairly quickly, after which the simulation eventually crashes)
>>> 
>>> Hi Vassili,
>>> 
>>> this error usually comes up when the arguments to umfpack_numeric() are not numbers (as would happen, for instance, if the metric had become singular). Could you check that your 3+1 variables are finite? Thorn NaNChecker could help you do that.
>> 
>> If this turns out to be the problem, we should probably check for this in AHFinderDirect and give an error message.  Something like "The metric has become singular at [x,y,z]; unable to compute horizon and aborting the run".
> 
> Nah: "unable to compute horizon; continuing the run".

This would be consistent with the behaviour when it is unable to find the horizon for other reasons, so I agree.  Though it might be good to have a "strict" parameter for each horizon which insists that it must be found when it is searched for, or the run stops.  Not sure what to do about horizons disappearing at a merger though.

-- 
Ian Hinder
http://numrel.aei.mpg.de/people/hinder