[Imstk-developers] Mesh load factor

Fri Sep 29 17:11:47 EDT 2017

Ok, now that makes more sense! I'll make an MR for this soon since this is pretty simple.
________________________________
From: Sreekanth Arikatla [sreekanth.arikatla at kitware.com]
Sent: Friday, September 29, 2017 4:50 PM
To: Milef, Nicholas Boris
Cc: imstk-developers at imstk.org
Subject: Re: [Imstk-developers] Mesh load factor

Sorry, all of my answers were assuming that preallocation to upper limit was not the strategy. If one preallocates to the limit load factor makes sense. I think it makes sense to set the upper limit to a certain factor of the original size.

On Fri, Sep 29, 2017 at 4:30 PM, Milef, Nicholas Boris <milefn at rpi.edu<mailto:milefn at rpi.edu>> wrote:
Sorry, I didn't mean in mechanical or aeronautics. I was more pointing towards its use in CS, electrical, and passenger load factor. But if it's confusing we can choose a new name (e.g., maxStorageFactor).

The limit doesn't change with each resize if that's what you're saying. You specify an upper limit (either an absolute limit or a scale from the original geometry, not sure which one is more appropriate). Then once you reach that limit, you get a warning and no more vertices can be added.
________________________________
From: Sreekanth Arikatla [sreekanth.arikatla at kitware.com<mailto:sreekanth.arikatla at kitware.com>]
Sent: Friday, September 29, 2017 4:13 PM

To: Milef, Nicholas Boris
Cc: imstk-developers at imstk.org<mailto:imstk-developers at imstk.org>
Subject: Re: [Imstk-developers] Mesh load factor

I know what load factor means in mech/aero fields. There is a common theme to the definition of load factor in different fields and that doesn't seem to be applicable here. 'Load factor' definition in comp sci there seems to be in the context of hash tables, which seems apt.

What I mean is if you end up with 20k verts from 10k with resizing 100 times is different from that with resizing once. To overcome all of this one generally, does a preallocation with the upper limit of the expected size.

On Fri, Sep 29, 2017 at 4:01 PM, Milef, Nicholas Boris <milefn at rpi.edu<mailto:milefn at rpi.edu>> wrote:
What do you mean by this?
________________________________
From: Sreekanth Arikatla [sreekanth.arikatla at kitware.com<mailto:sreekanth.arikatla at kitware.com>]
Sent: Friday, September 29, 2017 4:00 PM

To: Milef, Nicholas Boris
Cc: imstk-developers at imstk.org<mailto:imstk-developers at imstk.org>
Subject: Re: [Imstk-developers] Mesh load factor

The number of different instances the resizing happened also influences the fragmentation.

On Fri, Sep 29, 2017 at 3:56 PM, Sreekanth Arikatla <sreekanth.arikatla at kitware.com<mailto:sreekanth.arikatla at kitware.com>> wrote:
sounds good. Can we call it 'resizing factor' then?

On Fri, Sep 29, 2017 at 3:54 PM, Milef, Nicholas Boris <milefn at rpi.edu<mailto:milefn at rpi.edu>> wrote:
Well, I was thinking of just having a limit. So if you load a mesh that's 10k vertices, and it's load factor is 2x, then you can't add more than 20k vertices.
________________________________
From: Sreekanth Arikatla [sreekanth.arikatla at kitware.com<mailto:sreekanth.arikatla at kitware.com>]
Sent: Friday, September 29, 2017 3:50 PM
To: Milef, Nicholas Boris
Cc: imstk-developers at imstk.org<mailto:imstk-developers at imstk.org>
Subject: Re: [Imstk-developers] Mesh load factor

Can you explain a bit more? Do you plan to do a fresh allocation when resizing mesh beyond a certain size to prevent fragmentation?

On Fri, Sep 29, 2017 at 1:56 PM, Milef, Nicholas Boris <milefn at rpi.edu<mailto:milefn at rpi.edu>> wrote:
What do you all think about a mesh load factor? I would like to add one to prevent a range of allocations issues (fragmentation, invalid allocations, etc.) when resizing a mesh. This would be able to be customized by the user of course.

_______________________________________________
Imstk-developers mailing list
Imstk-developers at imstk.org<mailto:Imstk-developers at imstk.org>
http://public.kitware.com/mailman/listinfo/imstk-developers

--
Sreekanth Arikatla, Ph.D.,
Senior R&D Engineer,
Kitware, Inc.<http://www.kitware.com>, Carrboro, NC.

--
Sreekanth Arikatla, Ph.D.,
Senior R&D Engineer,
Kitware, Inc.<http://www.kitware.com>, Carrboro, NC.

--
Sreekanth Arikatla, Ph.D.,
Senior R&D Engineer,
Kitware, Inc.<http://www.kitware.com>, Carrboro, NC.

--
Sreekanth Arikatla, Ph.D.,
Senior R&D Engineer,
Kitware, Inc.<http://www.kitware.com>, Carrboro, NC.

--
Sreekanth Arikatla, Ph.D.,
Senior R&D Engineer,
Kitware, Inc.<http://www.kitware.com>, Carrboro, NC.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/imstk-developers/attachments/20170929/0ed82690/attachment.html>