• Sign In
  1. Issues List

Issue 630: rdotp uses Statistics.Angle() when it should be using NormalizedContrastAngleSqrt()

issues
Status:closed
Assigned To:Guest
Type:Defect
Area:Skyline
Priority:2
Milestone:21.2
Opened:2019-01-27 12:58 by Brendan MacLean
Changed:2022-02-23 17:11 by Brendan MacLean
Resolved:2021-11-29 13:46 by Rita Chupalov
Resolution:Fixed
Closed:2022-02-23 17:11 by Brendan MacLean
2019-01-27 12:58 Brendan MacLean
Title»rdotp uses Statistics.Angle() when it should be using NormalizedContrastAngleSqrt()
Assigned ToGuest»Nick Shulman
Type»Defect
Area»Skyline
Priority»2
Milestone»4.3
This issue got reported and analyzed on this support request:

https://skyline.ms/announcements/home/support/thread.view?rowId=38753

We need to be consistent, and Angle() anyway is too insensitive to change with small numbers of values. If we store these values in the Skyline document, then we may need to convert them, as I did for dotp and idopt when we switched from AngleSqrt() to NormalizedContrastAngleSqrt().
2019-06-10 12:45 Brendan MacLean
Assigned ToNick Shulman»Brendan MacLean
Milestone4.3»19.2
Need to change this.
2021-10-25 13:42 Brendan MacLean
Assigned ToBrendan MacLean»ritach
Notify»Nick Shulman
Milestone19.2»21.2
Since, you are doing so much work with rdotp display this release, I think we really should fix this issue, despite the fact that it will change the values. The values currently are inflated from what they would be if we used NormalizedContrastAngleSqrt(), which we use for dotp and idotp. Our naming makes them seem comparable, but they really are not, which is unfortunate.
2021-11-29 13:46 Rita Chupalov
resolve as Fixed
Statusopen»resolved
2022-02-23 17:11 Brendan MacLean
close
Statusresolved»closed
Assigned Toritach»Guest