Re: [racket-dev] Plot Contour Intervals Question (Bug?)

2012-08-04 Thread Neil Toronto

On 08/03/2012 11:59 AM, Doug Williams wrote:

Here is a simple contour-interval example that I don't understand.
[We are changing out some uses of shade in older code when I came
across something I simplified to this.]

Consider the following simple program.

(require plot)

(plot (contour-intervals (lambda (x y) (define z (- x y)) (cond (( z
-1) -1) (( z 1) 1) (else z))) -2 2 -2 2))

Whose output is attached.

Why is the lower white triangle not colored red - they definitely
have a value of 1? It's like those points are completely ignored.


It's definitely an error. Thanks for reporting it and providing a small 
test case.


I'll fix it today and get it into the release, so if you're planning to 
upgrade, you can proceed as if it works.


Neil ⊥
_
 Racket Developers list:
 http://lists.racket-lang.org/dev


Re: [racket-dev] Plot Contour Intervals Question (Bug?)

2012-08-04 Thread Neil Toronto

On 08/03/2012 11:59 AM, Doug Williams wrote:

(require plot)

(plot (contour-intervals
(lambda (x y)
  (define z (- x y))
  (cond (( z -1) -1)
(( z 1) 1)
(else z)))
-2 2 -2 2))

...

Why is the lower white triangle not colored red - they definitely have
a value of 1? It's like those points are completely ignored.


Turns out it was an error in the marching squares (intervals) 
implementation: a `' should have been a `=' to make an upper test 
symmetric with the corresponding lower test.


I've pushed the fix. Thanks again!

Neil ⊥

_
 Racket Developers list:
 http://lists.racket-lang.org/dev


Re: [racket-dev] Plot Contour Intervals Question (Bug?)

2012-08-04 Thread Doug Williams
Thanks, Neil.

On Saturday, August 4, 2012 x-apple-data-detectors://32, Neil Toronto
wrote:

 On 08/03/2012 11:59 AM, Doug Williams wrote:

 (require plot)

 (plot (contour-intervals
 (lambda (x y)
   (define z (- x y))
   (cond (( z -1) -1)
 (( z 1) 1)
 (else z)))
 -2 2 -2 2))

 ...

 Why is the lower white triangle not colored red - they definitely have
 a value of 1? It's like those points are completely ignored.


 Turns out it was an error in the marching squares (intervals)
 implementation: a `' should have been a `=' to make an upper test
 symmetric with the corresponding lower test.

 I've pushed the fix. Thanks again!

 Neil ⊥

 _
  Racket Developers list:
  http://lists.racket-lang.org/**dev http://lists.racket-lang.org/dev

_
  Racket Developers list:
  http://lists.racket-lang.org/dev