[yt-users] RAMSES MHD visualisation weirdness

[Sam Geen]

OK, looks like the problem was on my end - using the default YT field 
assignation works fine. I think the problem is that I assumed that YT 
adds dummy field names up to nvar, which is true, but only if it's 
already assigning automatic field names - it doesn't do this if you give 
it a field list by hand, which I had. Anyway, it works now, thanks for 
your advice.

On 11/02/2014 22:57, Sam Geen wrote:
> Thanks Romain - this ordering should already be implemented in the 
> frontend. I've just checked the code and RAMSES MHD does indeed add 3 
> to nvar before outputting - in this case I have no idea what caused 
> the bug in the previous outputs, unless an older version doesn't do 
> this. Will take another look tomorrow.
>
> Sam
>
> On 11/02/14 22:31, Romain Teyssier wrote:
>> Hi all,
>>
>> The RAMSES MHD output is indeed different than the hydro outputs.
>>
>> What is stored is in order:
>> 1- density
>> 2- x-velocity
>> 3- x-velocity
>> 4- x-velocity
>> 5- Left Bx field
>> 6- Left By field
>> 7- Left Bz field
>> 8- Right Bx field
>> 9- Right By field
>> 10- Right Bz field
>> 11- Thermal pressure
>> 12 and beyond: Passive scalars
>>
>>
>> The left and right B fields means that in Ramses the magnetic field 
>> is face centred.
>> So for each cell, you have a left-face and right-face centred quantity.
>>
>> Romain
>>
>>
>>
>>
>> On 11 Feb 2014, at 19:27, Sam Geen <samgeen at astro.ox.ac.uk> wrote:
>>
>>> Hi Nathan,
>>>
>>> Ah, thanks, you just gave me an idea - the MHD code (for some 
>>> reason, I've no idea why) has 3 extra variables than NVAR, and the 
>>> RAMSES frontend currently assumes that the output is reliably 
>>> reporting the number of variables, so it could be getting confused. 
>>> I'll try that tomorrow; you're right that this could be causing the 
>>> visualisation to wig out. This could be a bug/feature that's worth 
>>> changing in RAMSES itself as it's quite confusing, but since it's 
>>> existed for a while now it might be simpler just to work around it 
>>> in YT somehow.
>>>
>>> I'll look at getting together some simple MHD test problems and 
>>> making some outputs, too - there's at least one included with RAMSES 
>>> so I'll try running that now and contact you once I'm done.
>>>
>>> Sam
>>>
>>> On 11/02/14 19:02, Nathan Goldbaum wrote:
>>>> Hi Sam,
>>>>
>>>> Can you share the script you're using to make that plot?  In
>>>> particular I'd like to see the call you use to load the dataset, since
>>>> the number of fields you request and their order can matter a lot for
>>>> the RAMSES data format.
>>>>
>>>> It would also be nice to have a public RAMSES MHD test dataset - there
>>>> doesn't appear to be one on yt-project.org. If you (or anyone else on
>>>> the list) has one to share, that should help in debugging RAMSES MHD
>>>> issues in the future. If so, contact me off-list and we can work out
>>>> how to share the data.
>>>>
>>>> -Nathan
>>>>
>>>>
>>>> On Tue, Feb 11, 2014 at 7:58 AM, Sam Geen <samgeen at astro.ox.ac.uk> 
>>>> wrote:
>>>>> Hi,
>>>>>
>>>>> I tried visualising a RAMSES MHD output in YT and got this weird 
>>>>> effect:
>>>>> http://i.imgur.com/abjRorm.png (this happens for slices and 
>>>>> projections)
>>>>>
>>>>> Any idea what could be causing this? Nearly-identical non-MHD 
>>>>> simulations
>>>>> seem to visualise fine, and it happened for every MHD output I've 
>>>>> looked at.
>>>>> I can give more info if you like, although I'm not sure what 
>>>>> diagnostics to
>>>>> check for, so pointers on how to hunt this down would be 
>>>>> appreciated as
>>>>> well! It's entirely possible it's a bug/side-effect of the MHD 
>>>>> solver,
>>>>> although I'm not sure what, and other visualisation routines (in
>>>>> IDL/Fortran) seem to work fine with some of the outputs I've tried 
>>>>> on it.
>>>>>
>>>>> Thanks,
>>>>>
>>>>> Sam
>>>>> _______________________________________________
>>>>> yt-users mailing list
>>>>> yt-users at lists.spacepope.org
>>>>> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>>>> _______________________________________________
>>>> yt-users mailing list
>>>> yt-users at lists.spacepope.org
>>>> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>>> _______________________________________________
>>> yt-users mailing list
>>> yt-users at lists.spacepope.org
>>> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>> _______________________________________________
>> yt-users mailing list
>> yt-users at lists.spacepope.org
>> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org
>
> _______________________________________________
> yt-users mailing list
> yt-users at lists.spacepope.org
> http://lists.spacepope.org/listinfo.cgi/yt-users-spacepope.org