| Re: Uncaught PARTICLE_TRACE exception [message #39571] |
Fri, 28 May 2004 08:23 |
Karl Schultz
Messages: 341
Registered: October 1999
|
Senior Member |
|
|
"Karl Schultz" <kschultz_no_spam@rsinc.com> wrote in message
news:10bcckpsje3st3e@corp.supernews.com...
>
> "Ryan" <rab209@hotmail.com> wrote in message
> news:f845cc9f.0405270943.1a586412@posting.google.com...
>> Hi all,
>>
>> I'm using the PARTICLE_TRACE procedure to simulate a particle flow
>> through a 3-D velocity field, which, in most cases, works well.
>>
>> However, on the rare occasion where the seeds don't go anywhere
>> (usually due to the velocity vectors immediately pushing the seeds out
>> of the simulation volume), the program crashes, and IDL tells me that
>>
>> % Array dimensions must be greater than 0.
>>
>> and it points me to the line where my PARTICLE_TRACE call is made.
>> When examining the output values after the crash, the "Verts" and
>> "Normals" arrays are fine, but the "Conn" parameter is undefined...
>>
>> Which makes me think that PARTICLE_TRACE needs to have at least one
>> connecting line for it to work w/o crashing... which is unfortunate,
>> since I can't .cont my way back afterwards, nor can I automatically
>> predict if that would happen beforehand by looking at the data.
>>
>> Any ideas on if this is what's really happening and/or a workaround?
>>
>> Thanks much,
>> Ryan
>
> You are right about it not working when there are no lines to return.
> This has already been found and fixed for IDL 6.1.
> In 6.1, when there are no lines to return, the conn array is returned with
> the contents [-1].
> I'm sorry that I can't think of a good workaround.
I don't know why, but sometimes the answer pops up in my head a day later.
Use a CATCH block:
pro pt
data = FLTARR(2, 500, 500)
data[1,*,*] = 100000
seeds = [[500,500]]
catch, err
if err ne 0 then begin
print, "In catch"
help, verts
help, conn
conn = [-1]
catch, /cancel
endif $
else $
PARTICLE_TRACE, data, seeds, verts, conn
help, verts
help, conn
end
|
|
|
|
| Re: Uncaught PARTICLE_TRACE exception [message #39584 is a reply to message #39571] |
Thu, 27 May 2004 11:25  |
Karl Schultz
Messages: 341
Registered: October 1999
|
Senior Member |
|
|
"Ryan" <rab209@hotmail.com> wrote in message
news:f845cc9f.0405270943.1a586412@posting.google.com...
> Hi all,
>
> I'm using the PARTICLE_TRACE procedure to simulate a particle flow
> through a 3-D velocity field, which, in most cases, works well.
>
> However, on the rare occasion where the seeds don't go anywhere
> (usually due to the velocity vectors immediately pushing the seeds out
> of the simulation volume), the program crashes, and IDL tells me that
>
> % Array dimensions must be greater than 0.
>
> and it points me to the line where my PARTICLE_TRACE call is made.
> When examining the output values after the crash, the "Verts" and
> "Normals" arrays are fine, but the "Conn" parameter is undefined...
>
> Which makes me think that PARTICLE_TRACE needs to have at least one
> connecting line for it to work w/o crashing... which is unfortunate,
> since I can't .cont my way back afterwards, nor can I automatically
> predict if that would happen beforehand by looking at the data.
>
> Any ideas on if this is what's really happening and/or a workaround?
>
> Thanks much,
> Ryan
You are right about it not working when there are no lines to return.
This has already been found and fixed for IDL 6.1.
In 6.1, when there are no lines to return, the conn array is returned with
the contents [-1].
I'm sorry that I can't think of a good workaround.
Karl
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
