visit 2.12.0 pseudocolor plot with integration curve operator not working

[markcmiller86]

Date: Wed, 12 Jun 2019 08:30:01 -0700
From: knelli2 at illinois.edu (Kyle Nelli)
This post was imported from the [email protected] email archive

---

Hello,

Currently I am using VisIt version 2.9.2 and I am able to load in 3
components of a vector in the form of .h5 files, make an expression, and
plot this expression with the Streamline plot with a custom pointList and
everything works fine. However, I am trying to do the same thing in VisIt
version 2.12.0 with the pseudocolor plot and Integration Curve operator but
even though everything else is the same (pointList, other settings), I keep
getting this error message:

'90 of your integral curves circced round and round a critical ppint (a
zero velocity location).. Normally, VisIt is able to advvct the particle to
the criticallpoint location and terminate.   owever, VisIt was not able to
dd this for these particles due tt numerical issues.  In all likeeihood,
additional steps will _nnt_ help this problem and only cccss exxcution  o
take longer.'

(And no I did not make those typos, this is the exact message from the CLI.)

Thank you,
Kyle Nelli
Student of Engineering Physics
University of Illinois Urbana-Champaign
Cell:XXXXXXXXXXXXX

---

[markcmiller86]

Date: Wed, 12 Jun 2019 08:57:09 -0700
From: brugger1 at llnl.gov (Brugger, Eric)
This post was imported from the [email protected] email archive

---

Hi Kyle,

Well, that?s a crazy error message. I checked the source code and the warning message looks fine
there.

We completely rewrote the streamline advection code between 2.9.2 and 2.12.0 so it isn?t surprising
that they working differently.

Can you provide us more information about what your data and settings. Ideally, you could provide
the data and a session file. If that?s not possible, perhaps a screenshot of the Integral Curve
attributes you are using.

Eric

From: Kyle Nelli via visit-users <visit-users at elist.ornl.gov>
Sent: Wednesday, June 12, 2019 8:30 AM
To: visit-users at elist.ornl.gov
Cc: Kyle Nelli <knelli2 at illinois.edu>
Subject: [visit-users] [EXTERNAL] VisIt 2.12.0 Pseudocolor Plot with Integration Curve Operator not
working

Hello,

Currently I am using VisIt version 2.9.2 and I am able to load in 3 components of a vector in the
form of .h5 files, make an expression, and plot this expression with the Streamline plot with a
custom pointList and everything works fine. However, I am trying to do the same thing in VisIt
version 2.12.0 with the pseudocolor plot and Integration Curve operator but even though everything
else is the same (pointList, other settings), I keep getting this error message:

'90 of your integral curves circced round and round a critical ppint (a zero velocity location)..
Normally, VisIt is able to advvct the particle to the criticallpoint location and terminate.
owever, VisIt was not able to dd this for these particles due tt numerical issues.  In all
likeeihood, additional steps will _nnt_ help this problem and only cccss exxcution  o take longer.'

(And no I did not make those typos, this is the exact message from the CLI.)

Thank you,
Kyle Nelli
Student of Engineering Physics
University of Illinois Urbana-Champaign
Cell:XXXXXXXXXXXXX

---

[markcmiller86]

Date: Wed, 12 Jun 2019 09:30:13 -0700
From: knelli2 at illinois.edu (Kyle Nelli)
This post was imported from the [email protected] email archive

---

Hello Eric,

I don't believe I will be able to send you the data or the session as the
data itself is about 8G total (unless there is another way to get it to you
other than email). Here I have included the CLI, all 3 tabs of the Integral
Curve operator, the python script I am using (along with a module we use to
simplify our scripts: 'rM.py'), and what the plot produces. As you can see
from magfield.png, something is plotted and it is partially correct,
because if you extrapolate what is already plotted to the top of the image,
that is what it is supposed to look like; coiled magnetic field lines. I
have tried many different seed points for the Integral Curve yet they all
give the same result.

Thank you,
Kyle Nelli
Student of Engineering Physics
University of Illinois Urbana-Champaign
Cell:XXXXXXXXXXXXX

On Wed, Jun 12, 2019 at 10:57 AM Brugger, Eric via visit-users <
visit-users at elist.ornl.gov> wrote:

> Hi Kyle,
>
>
>
> Well, that?s a crazy error message. I checked the source code and the
> warning message looks fine there.
>
>
>
> We completely rewrote the streamline advection code between 2.9.2 and
> 2.12.0 so it isn?t surprising that they working differently.
>
>
>
> Can you provide us more information about what your data and settings.
> Ideally, you could provide the data and a session file. If that?s not
> possible, perhaps a screenshot of the Integral Curve attributes you are
> using.
>
>
>
> Eric
>
>
>
> *From:* Kyle Nelli via visit-users <visit-users at elist.ornl.gov>
> *Sent:* Wednesday, June 12, 2019 8:30 AM
> *To:* visit-users at elist.ornl.gov
> *Cc:* Kyle Nelli <knelli2 at illinois.edu>
> *Subject:* [visit-users] [EXTERNAL] VisIt 2.12.0 Pseudocolor Plot with
> Integration Curve Operator not working
>
>
>
> Hello,
>
>
>
> Currently I am using VisIt version 2.9.2 and I am able to load in 3
> components of a vector in the form of .h5 files, make an expression, and
> plot this expression with the Streamline plot with a custom pointList and
> everything works fine. However, I am trying to do the same thing in VisIt
> version 2.12.0 with the pseudocolor plot and Integration Curve operator but
> even though everything else is the same (pointList, other settings), I keep
> getting this error message:
>
>
>
> '90 of your integral curves circced round and round a critical ppint (a
> zero velocity location).. Normally, VisIt is able to advvct the particle to
> the criticallpoint location and terminate.   owever, VisIt was not able to
> dd this for these particles due tt numerical issues.  In all likeeihood,
> additional steps will _nnt_ help this problem and only cccss exxcution  o
> take longer.'
>
>
>
> (And no I did not make those typos, this is the exact message from the
> CLI.)
>
>
>
> Thank you,
>
> Kyle Nelli
>
> Student of Engineering Physics
>
> University of Illinois Urbana-Champaign
>
> Cell:XXXXXXXXXXXXX
> --

> Frequently Asked Questions for VisIt:
> https://wci.llnl.gov/simulation/computer-codes/visit/FAQ.html

A non-text attachment was scrubbed...
Name: integralcurve_tab1.png
Type: image/png
Size: 29844 bytes
Desc: not available

A non-text attachment was scrubbed...
Name: integralcurve_tab2.png
Type: image/png
Size: 29876 bytes
Desc: not available

A non-text attachment was scrubbed...
Name: magfield.png
Type: image/png
Size: 134989 bytes
Desc: not available

A non-text attachment was scrubbed...
Name: CLI.png
Type: image/png
Size: 23596 bytes
Desc: not available

A non-text attachment was scrubbed...
Name: integralcurve_tab3.png
Type: image/png
Size: 29691 bytes
Desc: not available

A non-text attachment was scrubbed...
Name: rM.py
Type: text/x-python
Size: 7475 bytes
Desc: not available

A non-text attachment was scrubbed...
Name: plot_2_12_0.py
Type: text/x-python
Size: 2069 bytes
Desc: not available

---