Re: Regarding TTK's morse smale complex

[Julien Tierny]

Dear Ritesh,

thanks for your email.

> I have a point cloud which I am converting to unstructured grid and each vertex in the grid contains the scalar values.
I confirm that TTK can build the Morse-Smale complex for this kind of input.
Please don't hesitate to share your data to the mailing list if you're running into further issues.

Best regards,
--
Dr Julien Tierny
CNRS Researcher
Sorbonne Universite
http://lip6.fr/Julien.Tierny

On Monday, November 7, 2022 4:36:22 AM CET Ritesh Sharma wrote:
> Hi,
>
> I am trying to use TTK's morse smale complex for a piecewise constant scalar field i.e, a distance field.
> I have a point cloud which I am converting to unstructured grid and each vertex in the grid contains the scalar values.
> I would like to know if the TTK's morse smale complex can be computed for a piecewise constant scalar field i.,e a distance field?
>
> Regards,
> Ritesh
>
>
> PhD Candidate, EECS
> Computer Graphics Lab, School of Engineering,
> University of California, Merced, California
> Homepage:https://sharmrit.github.io/Homepage/
>
>

[Ritesh Sharma]

Hello Dr Tierny,

I have a point cloud in the form of a text file(scalar_data_2.txt) which is basically a file containing x,yz and distance columns. I am converting the data into unstructured grid(unstructuredgrid.vtu) using the python script(vtkfilegenerator_Scalar.py). Once I have unstructuredgrid.vtu file, I open the data in the Paraview and perform Delaunay 2D triangulation and then compute morse smale complex. The result, I am getting isn't able to compute all the critical points using morse smale complex(shown in obtained_result.PNG). 

obtained_result.PNG

unstructuredgrid 1.vtu

vtkfilegenerator_Scalar.py

scalar_data 2.txt

Regards,

Ritesh

PhD Candidate, EECS

Computer Graphics Lab, School of Engineering,

University of California, Merced, California

Homepage:https://sharmrit.github.io/Homepage/

---

From: Julien Tierny <julien...@sorbonne-universite.fr>
Sent: Monday, November 7, 2022 9:07 PM
To: ttk-...@googlegroups.com <ttk-...@googlegroups.com>; Ritesh Sharma <rsha...@ucmerced.edu>
Subject: Re: Regarding TTK's morse smale complex

 

[Julien Tierny]

hi ritesh,

thanks for your email.

> I have a point cloud in the form of a text file(scalar_data_2.txt) which is basically a file containing x,yz and distance columns.

understood.

> The result, I am getting isn't able to compute all the critical points using morse smale complex(shown in obtained_result.PNG).

I have applied a similar pipeline (attached PVSM file) and the result is on par with expectations.

please consider the screenshot "ritesh_simplified.png".
on the left of the screen, I used discrete morse theory (with the "morse smale complex" filter) to extract the critical points.
on the right, I used the piecewise linear formalism to extract critical points (with the "scalar field critical points" filter).

the difference between the two outputs is essentially on the boundary of the domain (in particular critical points cannot occur on the boundary in discrete morse theory). other than that, the output is correct.

of course, I used some level of persistence-based simplification to get rid of small oscillations on flat plateaus (the original analysis is shown in "ritesh_raw.png").

so for me, it's all good.

I hope this helps.

best regards,

--
Dr Julien Tierny
CNRS Researcher
Sorbonne Universite
http://lip6.fr/Julien.Tierny

On Tuesday, November 8, 2022 6:30:19 AM CET Ritesh Sharma wrote:
> Hello Dr Tierny,
>
> I have a point cloud in the form of a text file(scalar_data_2.txt) which is basically a file containing x,yz and distance columns.
> I am converting the data into unstructured grid(unstructuredgrid.vtu) using the python script(vtkfilegenerator_Scalar.py). Once I have unstructuredgrid.vtu file, I open the data in the Paraview and perform Delaunay 2D triangulation and then compute morse smale complex.
> The result, I am getting isn't able to compute all the critical points using morse smale complex(shown in obtained_result.PNG).
>

> [https://res-geo.cdn.office.net/assets/mail/file-icon/png/photo_16x16.png]obtained_result.PNG<https://merced-my.sharepoint.com/:i:/g/personal/rsharma39_ucmerced_edu/EbMWFZ-kHY5HhbSdegyo7XsB55N_UdbDze93u8bZnjNdwg>[cid:4b25f82f-83b8-43cc-95f8-0a7411864895]
> [https://res-geo.cdn.office.net/assets/mail/file-icon/png/generic_16x16.png]unstructuredgrid 1.vtu<https://merced-my.sharepoint.com/:u:/g/personal/rsharma39_ucmerced_edu/ESUNumP4btRHhmpKLhXhkswBxyAIA4J6R7_zFguVHD501A>
> [https://res-geo.cdn.office.net/assets/mail/file-icon/png/generic_16x16.png]vtkfilegenerator_Scalar.py<https://merced-my.sharepoint.com/:u:/g/personal/rsharma39_ucmerced_edu/EUWMd5Goba1Pt7rGOCq3S8EBXs2BB3B3cBrkJxIuhAj9Yg>
> [https://res.cdn.office.net/assets/mail/file-icon/png/txt_16x16.png]scalar_data 2.txt<https://merced-my.sharepoint.com/:t:/g/personal/rsharma39_ucmerced_edu/ESnkAHrXj5NMlhgPo16dXTMB3GSm0CZY2390yi_6ZkYSpw>