MathGroup Archive 2001

[Date Index] [Thread Index] [Author Index]

Search the Archive

Graphics Bug (or anomaly anyway)

  • To: mathgroup at
  • Subject: [mg28303] Graphics Bug (or anomaly anyway)
  • From: aes <siegman at>
  • Date: Mon, 9 Apr 2001 02:58:03 -0400 (EDT)
  • Sender: owner-wri-mathgroup at

Try executing the following code, which creates a rectangular array of 
points extending from n=0 to n=nn horizontally and m=1 to m=nn 

         Print["nn = ", nn];
              Table[Point[{n, m}], {n, 0, nn}, {m, 1, nn}]]],  {nn, 5,8}]

Then click on the plots to show the bounding boxes.

For nn<=6 the bounding box is closely wrapped around the array, with the 
bottom line especially located just a little below the m=1 row.

For nn>=7 the bottom of the bounding box suddenly drops down one unit 
and is now located a little below where an m=0 row would be, if there 
were one.

If you make the array run only from n=1 and m=1 in both directions, the 
bounding box wraps tightly all around for nn<=6, but drops back by one 
unit on both left and bottom sides for nn>=7.

Try giving a larger PointSize to the points.  If the points are large 
enough the circular points on the outer boundary of the array get cut 
off by the bounding box on all 4 sides for nn<=6, but only on the top 
and two sides, not on the bottom, for n>=7.

Same thing happens with Circle or Disk -- except bounding box always 
expands enough so the circles or disks are never cut off.

Not a serious bug, maybe -- but does produce some minor hassles, e.g., 
if you're exporting a series of figures at increasing values of nn to go 
into another application, and counting on the bounding boxes to be all 
the same size.

[Mathematica 4.0, Mac PB G3, OS 8.6]

  • Prev by Date: I'm new
  • Next by Date: Re: Laplace Transforms of piecewise continuous functions
  • Previous by thread: I'm new
  • Next by thread: RE: Graphics Bug (or anomaly anyway)