Update isd_generate.py to support optional function to ensure that instrument pointing quaternions do not change sign

[thareUSGS]

Implement an optional function to ensure that instrument pointing quaternions do not change sign.

This is really more to continue the discussion under issue #603 . I understand that NAIF is supposed to return the quaternions with flipped signs, but this seems to be breaking our applications. ASP has been updated to account for this issue. This example function for isd_generate.py, following ASP's logic, could also be used to "clean" the ISD for other applications like SOCET GXP.

Licensing

This project is mostly composed of free and unencumbered software released into the public domain, and we are unlikely to accept contributions that are not also released into the public domain. Somewhere near the top of each file should have these words:

This work is free and unencumbered software released into the public domain. In jurisdictions that recognize copyright laws, the author or authors of this software dedicate any and all copyright interest in the software to the public domain.

• I dedicate any and all copyright interest in this software to the public domain. I make this dedication for the benefit of the public at large and to the detriment of my heirs and successors. I intend this dedication to be an overt act of relinquishment in perpetuity of all present and future rights to this software under copyright law.

[Kelvinrr]

Are there cases where you wouldn't want this on?

[thareUSGS]

Good question. I think @oleg-alexandrov would state this should always be run to fix for all images (that assumes I am doing this correctly). But perhaps we have a flag that turns this function off instead of on (implying we would always run it unless the flag stops it).

[jlaura]

Weighing in here with a +1 to @oleg-alexandrov's comment below. Always on.

[oleg-alexandrov]

This must run always. This is fix for quartertions having two solutions for any one rotation. In rare cases one may want the option of not doing this.

…

On Wed, Sep 4, 2024, 11:56 PM Trent Hare ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In ale/isd_generate.py <#612 (comment)>: > + parser.add_argument( + "-f", "--fix_quaternion", + action="store_true", + help="There are rare cases when a stream of quaternion flip their signs. This option will " + "fix those to be consistent. This takes after a NASA ASP function to do the same. " + ) Good question. I think @oleg-alexandrov <https://github.com/oleg-alexandrov> would state this should always be run to fix for all images (that assumes I am doing this correctly). But perhaps. we have a flag that turns this function off instead of on. — Reply to this email directly, view it on GitHub <#612 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAKDU3GH274J7UL3VRIR2MDZU5XYLAVCNFSM6AAAAABM4WMXRSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDEOBRGMYTMNJQGA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[oleg-alexandrov]

To add, this fix must be deep in ALE, in the linescan driver. That some quarternions flip sign is a limitation of the math. No tools will break if the sign is regularized. Nothing is gained by allowing a flip, as any one quarterion is equally good with any sign. On Thu, Sep 5, 2024, 3:04 AM Oleg Alexandrov ***@***.***> wrote:

…

This must run always. This is fix for quartertions having two solutions for any one rotation. In rare cases one may want the option of not doing this. On Wed, Sep 4, 2024, 11:56 PM Trent Hare ***@***.***> wrote: > ***@***.**** commented on this pull request. > ------------------------------ > > In ale/isd_generate.py > <#612 (comment)>: > > > + parser.add_argument( > + "-f", "--fix_quaternion", > + action="store_true", > + help="There are rare cases when a stream of quaternion flip their signs. This option will " > + "fix those to be consistent. This takes after a NASA ASP function to do the same. " > + ) > > Good question. I think @oleg-alexandrov > <https://github.com/oleg-alexandrov> would state this should always be > run to fix for all images (that assumes I am doing this correctly). But > perhaps. we have a flag that turns this function off instead of on. > > — > Reply to this email directly, view it on GitHub > <#612 (comment)>, or > unsubscribe > <https://github.com/notifications/unsubscribe-auth/AAKDU3GH274J7UL3VRIR2MDZU5XYLAVCNFSM6AAAAABM4WMXRSVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDEOBRGMYTMNJQGA> > . > You are receiving this because you were mentioned.Message ID: > ***@***.***> >

[Kelvinrr]

We need to create a new PR that moves this code into the ALE mixing