Hi TTK:
I am able to reproduce the results in “Morse molecule demo”. However, when I tried to implement morse smale complex on my own 3D scalar data. There are still a lot of redundant and undesired minimum points. Even when I reduced the persistent pair down to “1 pair”, those minimum points are still there, just like the second figure below. I don’t have any problem with 2D, but results for 3D data are confusing.
Is there any way to address this problem?
Scalar field
Critical points after reduce persistent pair down to 1 pair
Critical points with scalar field overlapped
Thanks,
Lee, Shao-Chun
Graduate Research Assistant
Beckman Institute for Advanced Science and Technology
University of Illinois at Urbana-Champaign
3614 Beckman Institute
405 N. Mathews Ave.
Urbana, IL 61801
217 819 8722 | sc...@illinois.edu
Hi Jonas Lukasczyk:
Thank you for your prompt reply. Attachment includes log file and screenshot of the pipeline. PT is “Threshold” for persistence.
Best,
Lee, Shao-Chun
Graduate Research Assistant
Beckman Institute for Advanced Science and Technology
University of Illinois at Urbana-Champaign
3614 Beckman Institute
405 N. Mathews Ave.
Urbana, IL 61801
217 819 8722 | sc...@illinois.edu
> To view this discussion on the web visit https://urldefense.com/v3/__https://groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHBiJzQ4m$ (https://urldefense.com/v3/__https://groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com?utm_medium=email&utm_source=footer__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHLuhJf77$ ).
--
You received this message because you are subscribed to the Google Groups "ttk-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ttk-users+...@googlegroups.com.
Hi,
I am currently using TTK.0.9.7. If that’s the case, I will re-install TTK with the latest version. But it would be great if you could try this data.
Thanks,
Lee, Shao-Chun
Graduate Research Assistant
Beckman Institute for Advanced Science and Technology
University of Illinois at Urbana-Champaign
3614 Beckman Institute
405 N. Mathews Ave.
Urbana, IL 61801
217 819 8722 | sc...@illinois.edu
--
You received this message because you are subscribed to the Google Groups "ttk-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ttk-users+...@googlegroups.com.
thanks for your email and sorry for the late reply.
what you observe is... normal :)
see the attached screenshot `latent_post_lts.png` where I reproduced your pipeline with a persistence threshold set to 0.05.
there, the spheres are color-coded as follows:
- dark blue: minimum (there's only one of them)
- dark green: maxima (there's only 10 of them, lts did its job)
- light blue: 1-saddles
- light green: 2-saddles
what happens for this dataset is that there are thousands of 1-saddle/2-saddle pairs due to the low-scale noise in the data.
these correspond to tiny tunnels that appear in the isosurfaces.
unfortunately, lts (as well as the legacy simplification backend) does not support the removal of 1-saddle/2-saddle pair.
there is a workaround though, which is an experimental feature at the moment (it may get a little buggy).
when computing the Morse-Smale complex, check the box `Return Saddle Connectors`. then a new text box appears, labeled `Saddle Connectors Persistence Threshold`. there enter the same persistence threshold as previously (here 0.05).
the result is shown in the screenshot `latent_clean`.
there you will see that only the saddle pairs involved in large tunnels (such as the one pointed by my cursor) are maintained.
what this experimental feature does is that it post-processes the Morse-Smale complex to remove pairs corresponding to small tunnels. note that it does not simplify your input data as `TopologicalSimplification` does unfortunately. so only the Morse-Smale complex will benefit from this simplification (and not the other tools available in TTK).
note that since your data looked very noisy in the lower range, I applied a very mild smoothing as a pre-processing (only 1 iteration was sufficient).
the pipeline is attached to the email `latent.pvsm`.
note that you'll need to change the path the input accordingly (File, Load State File, in the resulting dialog window, select "Choose File Names" and select your input file). also, I used the dev branch of ttk and paraview 5.9 but all of this works in principle with previous versions.
just to be safe though, as jonas suggested, I'd really invite you to upgrade to the latest stable version (0.9.9). this version has more features and is significantly faster too. its installation has been made significantly easier too (https://topology-tool-kit.github.io/installation.html)
I hope this helps.
please don't hesitate to let us know if you have further questions.
cheers,
--
Dr Julien Tierny
CNRS Researcher
Sorbonne Universite
http://lip6.fr/Julien.Tierny
Dear Jonas:
I really appreciate your help. This is exactly what I am looking for. I guess the experimental feature you mentioned only appears in the latest version, right? Because I cannot find those options (`Return Saddle Connectors`) in the properties of TTKMorseSmalComplex pipeline.
Thanks,
Lee, Shao-Chun
Graduate Research Assistant
Beckman Institute for Advanced Science and Technology
University of Illinois at Urbana-Champaign
3614 Beckman Institute
405 N. Mathews Ave.
Urbana, IL 61801
217 819 8722 | sc...@illinois.edu
From: Julien Tierny
Sent: Tuesday, February 9, 2021 4:52 AM
To: ttk-...@googlegroups.com;
Lee, Shao-Chun
Cc: Jonas Lukasczyk
Subject: Re: [ttk-users] Morse Smale Complex on 3D scalarfield
hi guys,
thanks for your email and sorry for the late reply.
what you observe is... normal :)
see the attached screenshot `latent_post_lts.png` where I reproduced your pipeline with a persistence threshold set to 0.05.
there, the spheres are color-coded as follows:
- dark blue: minimum (there's only one of them)
- dark green: maxima (there's only 10 of them, lts did its job)
- light blue: 1-saddles
- light green: 2-saddles
what happens for this dataset is that there are thousands of 1-saddle/2-saddle pairs due to the low-scale noise in the data.
these correspond to tiny tunnels that appear in the isosurfaces.
unfortunately, lts (as well as the legacy simplification backend) does not support the removal of 1-saddle/2-saddle pair.
there is a workaround though, which is an experimental feature at the moment (it may get a little buggy).
when computing the Morse-Smale complex, check the box `Return Saddle Connectors`. then a new text box appears, labeled `Saddle Connectors Persistence Threshold`. there enter the same persistence threshold as previously (here 0.05).
the result is shown in the screenshot `latent_clean`.
there you will see that only the saddle pairs involved in large tunnels (such as the one pointed by my cursor) are maintained.
what this experimental feature does is that it post-processes the Morse-Smale complex to remove pairs corresponding to small tunnels. note that it does not simplify your input data as `TopologicalSimplification` does unfortunately. so only the Morse-Smale complex
will benefit from this simplification (and not the other tools available in TTK).
note that since your data looked very noisy in the lower range, I applied a very mild smoothing as a pre-processing (only 1 iteration was sufficient).
the pipeline is attached to the email `latent.pvsm`.
note that you'll need to change the path the input accordingly (File, Load State File, in the resulting dialog window, select "Choose File Names" and select your input file). also, I used the dev branch of ttk and paraview 5.9 but all of this works in principle
with previous versions.
just to be safe though, as jonas suggested, I'd really invite you to upgrade to the latest stable version (0.9.9). this version has more features and is significantly faster too. its installation has been made significantly easier too (https://urldefense.com/v3/__https://topology-tool-kit.github.io/installation.html__;!!DZ3fjg!pVh04kcZrU7GM8ZE0wQf85UV14JDRF67yNxsOd4eJUDtiqWtuSgpEpeNWHiXbbgA$ )
I hope this helps.
please don't hesitate to let us know if you have further questions.
cheers,
--
Dr Julien Tierny
CNRS Researcher
Sorbonne Universite
> I guess the experimental feature you mentioned only appears in the latest version, right? Because I cannot find those options (`Return Saddle Connectors`) in the properties of TTKMorseSmalComplex pipeline.
the feature appears in the "advanced" mode, which you can enable by clicking on the small button with a gear-like icon, right next to the search bar of the properties panel. (actually, you can type "return" in that search bar as well to find the check box I mentioned).
but again, I strongly invite you to upgrade your version of ttk anyways :)
cheers,
--
Dr Julien Tierny
CNRS Researcher
Sorbonne Universite
On Tuesday, 9 February 2021 16:29:07 CET Lee, Shao-Chun wrote:
> Dear Jonas:
>
> I really appreciate your help. This is exactly what I am looking for.
> I guess the experimental feature you mentioned only appears in the latest version, right? Because I cannot find those options (`Return Saddle Connectors`) in the properties of TTKMorseSmalComplex pipeline.
>
> Thanks,
> Lee, Shao-Chun
> Graduate Research Assistant
>
> Beckman Institute for Advanced Science and Technology
> University of Illinois at Urbana-Champaign
> 3614 Beckman Institute
> 405 N. Mathews Ave.
> Urbana, IL 61801
> 217 819 8722 | sc...@illinois.edu
>
> https://urldefense.com/v3/__http://lip6.fr/Julien.Tierny__;!!DZ3fjg!pVh04kcZrU7GM8ZE0wQf85UV14JDRF67yNxsOd4eJUDtiqWtuSgpEpeNWMOglr0U$<https://urldefense.com/v3/__http:/lip6.fr/Julien.Tierny__;!!DZ3fjg!pVh04kcZrU7GM8ZE0wQf85UV14JDRF67yNxsOd4eJUDtiqWtuSgpEpeNWMOglr0U$>
> > > > > To view this discussion on the web visit https://urldefense.com/v3/__https://groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHBiJzQ4m$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHBiJzQ4m$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHBiJzQ4m$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHBiJzQ4m$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHBiJzQ4m$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHBiJzQ4m$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHBiJzQ4m$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHBiJzQ4m$>>> (https://urldefense.com/v3/__https://groups.google.com/d/msgid/ttk-users/CH2PR11MB4486862A5427FC1BFA5ED1C4F1B19*40CH2PR11MB4486.namprd11.prod.outlook.com?utm_medium=email&utm_source=footer__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHLuhJf77$ ).
> > > >
> > > > --
> > > > You received this message because you are subscribed to the Google Groups "ttk-users" group.
> > > > To unsubscribe from this group and stop receiving emails from it, send an email to ttk-users+...@googlegroups.com.
> > > > To view this discussion on the web visit https://urldefense.com/v3/__https://groups.google.com/d/msgid/ttk-users/306871662.607807.1612652784399*40webmail.strato.de__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHKsFlJAQ$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/306871662.607807.1612652784399*40webmail.strato.de__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHKsFlJAQ$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/306871662.607807.1612652784399*40webmail.strato.de__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHKsFlJAQ$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/306871662.607807.1612652784399*40webmail.strato.de__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHKsFlJAQ$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/306871662.607807.1612652784399*40webmail.strato.de__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHKsFlJAQ$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/306871662.607807.1612652784399*40webmail.strato.de__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHKsFlJAQ$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/306871662.607807.1612652784399*40webmail.strato.de__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHKsFlJAQ$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/306871662.607807.1612652784399*40webmail.strato.de__;JQ!!DZ3fjg!upW1Slv3u9Iplaoj2ZSZ_BsSpFDrD7OUlSN9iTPm9PFqHdfnaDceeZiUHKsFlJAQ$>>> .
> > > >
> > > > --
> > > > You received this message because you are subscribed to the Google Groups "ttk-users" group.
> > > > To unsubscribe from this group and stop receiving emails from it, send an email to ttk-users+...@googlegroups.com.
> > > > To view this discussion on the web visit https://urldefense.com/v3/__https://groups.google.com/d/msgid/ttk-users/CH2PR11MB4486A83C7B8994F63FE3546AF1B09*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!pktv28eP_p5pQfxipalnz8wBtsTbRyg1psfyJHnvOkxqv4OEGH9DPSGknJhJKdYQ$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486A83C7B8994F63FE3546AF1B09*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!pktv28eP_p5pQfxipalnz8wBtsTbRyg1psfyJHnvOkxqv4OEGH9DPSGknJhJKdYQ$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486A83C7B8994F63FE3546AF1B09*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!pktv28eP_p5pQfxipalnz8wBtsTbRyg1psfyJHnvOkxqv4OEGH9DPSGknJhJKdYQ$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/CH2PR11MB4486A83C7B8994F63FE3546AF1B09*40CH2PR11MB4486.namprd11.prod.outlook.com__;JQ!!DZ3fjg!pktv28eP_p5pQfxipalnz8wBtsTbRyg1psfyJHnvOkxqv4OEGH9DPSGknJhJKdYQ$>> .
> > >
> > > --
> > > You received this message because you are subscribed to the Google Groups "ttk-users" group.
> > > To unsubscribe from this group and stop receiving emails from it, send an email to ttk-users+...@googlegroups.com.
> > > To view this discussion on the web visit https://urldefense.com/v3/__https://groups.google.com/d/msgid/ttk-users/1700477632.675223.1612787103803*40webmail.strato.de__;JQ!!DZ3fjg!pktv28eP_p5pQfxipalnz8wBtsTbRyg1psfyJHnvOkxqv4OEGH9DPSGknKA8QJAC$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/1700477632.675223.1612787103803*40webmail.strato.de__;JQ!!DZ3fjg!pktv28eP_p5pQfxipalnz8wBtsTbRyg1psfyJHnvOkxqv4OEGH9DPSGknKA8QJAC$<https://urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/1700477632.675223.1612787103803*40webmail.strato.de__;JQ!!DZ3fjg!pktv28eP_p5pQfxipalnz8wBtsTbRyg1psfyJHnvOkxqv4OEGH9DPSGknKA8QJAC$%3chttps:/urldefense.com/v3/__https:/groups.google.com/d/msgid/ttk-users/1700477632.675223.1612787103803*40webmail.strato.de__;JQ!!DZ3fjg!pktv28eP_p5pQfxipalnz8wBtsTbRyg1psfyJHnvOkxqv4OEGH9DPSGknKA8QJAC$>> .
> > >
> >
> >
>
>