Profile of sphere

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Profile of sphere

    Just out of curiosity. I have touched on this subject before. I have to report a Profile back to itself. I normally report it Least Square. I chose Vector Least Square and it came out better. What is difference on this setting actually than that of Least Square? From my understanding it is a result of better actual hits taken.
    Last edited by KIRBSTER269; 02-14-2017, 01:48 PM.
    (In Memory of my Loving wife, "Ronda" who I lost March 7, 2016. I love you baby.)
    They say "Nobody's Perfect." I must be Nobody.

  • #2
    1. Least Squares (default) - minimizes the average squared error of the fit among all of the features included in the best fit. This is the most common best fit type. The error value is the sum of the squared distances. Outliers don't have as bad an effect and direction of error plays no role in this fit.
    2. Vector (also known as Projected Least Squares) - also minimizes the average squared error among the features being fit; however, it first snaps (or projects) the points back onto the nominal vectors of the feature. Therefore, all error will be along the nominal vectors. The error value is the sum of the squared projected distances. This type of fit is typically used when a set of points is being fit to curves and / or surfaces. With this type of fit, PC-DMIS allows points to 'slide' along the surface but discourages them from pulling away from the surface.
      Example: If the Nominal point is: 1.1.1 with the vector: 0,0,1, and Measure values 4,2,0.95, then the measured data will be adjusted to 1,1,0.95, snapped to the vector of 0,0,1.
    From my understanding I should be using Vector Least square instead of Least Square.

    Correction I have been using Vector Least square.
    Last edited by KIRBSTER269; 02-14-2017, 03:18 PM. Reason: Correction
    (In Memory of my Loving wife, "Ronda" who I lost March 7, 2016. I love you baby.)
    They say "Nobody's Perfect." I must be Nobody.

    Comment


    • #3
      have you tried using vector least square? (kidding) in addition, there is vector min max in newer versions, but TP forum was closed before I could ask Rob why I got such a big difference in profile between
      vector LS and vector min max. Perhaps a guru can shine some light?

      Comment


      • #4
        I'm not at my machine but if I remember correctly the help files say that the least square is generally for planar features whereas the vector least square is for non flat features. I'll check when I get in tomorrow if I haven't been corrected or confirmed by then.

        Comment


        • KIRBSTER269
          KIRBSTER269 commented
          Editing a comment
          That's what I copied out of the help file on the 2nd post. 4th sentence in the vector part, states, for curved surfaces.

        • william
          william commented
          Editing a comment
          Ha, yep I see that now.

        • KIRBSTER269
          KIRBSTER269 commented
          Editing a comment
          That's quite alright william .I still expect you to throw your 2 cents in. I want as much feedback as I can. Makes me a little more sure of myself.

      Related Topics

      Collapse

      Working...
      X