Skip to main content
Question

domain violation

  • June 2, 2016
  • 2 replies
  • 24 views
gio
Contributor
Forum|alt.badge.img+15

When using the arithmitic editor i found it tends to often lead to domainviolation when using goniometic calculations:

 

2016-06-02 12:51:34| 67.4| 0.0|WARN |Number of Coordinates: 2 -- Coordinate Dimension: 2

 

2016-06-02 12:51:34| 67.4| 0.0|WARN |(95221.671876227498,464970.131574183)(95220.230171277391,464970.131574183)

 

2016-06-02 12:51:34| 67.4| 0.0|WARN |===========================================================================

 

2016-06-02 12:51:36| 69.7| 2.2|WARN |Hoek_in_graden_en_radialen._2_AttributeCreator_2: Failed to evaluate arithmetic expression '-1.#IND*(180*1.0/3.1415926535897932)'

 

2016-06-02 12:51:36| 69.7| 0.0|WARN |Hoek_in_graden_en_radialen._2_AttributeCreator_2: TCL Error Message:

 

2016-06-02 12:51:36| 69.7| 0.0|WARN | invalid character "#"

 

2016-06-02 12:51:36| 69.7| 0.0|WARN | in expression "-1.#IND*(180*1.0/3.1415926..."

 

2016-06-02 12:51:36| 69.7| 0.0|WARN |Hoek_in_graden_en_radialen._2_AttributeCreator_2: Failed to evaluate expression '-1.#IND*(180*1.0/3.1415926535897932)'

This bit in the above expression uising above vertices

@Evaluate((@Value(E_x)-@Value(S_x))/@sqrt(@pow(@Value(E_x)-@Value(S_x),2)))

has fme output -1.0000000000000027 (the table view shows value -1 b.t.w. so its a bit lying...lol)

wich leads to acos (-1.0000000000000027) wich is out of domain.

Where is should be

-1 (in fact when line is horizontal it is calculating A/SQRT(Pow(A) wich is A/A wich is 1.. child can do it..)

Fme has no pmin, pmax functions.

the expr tcl function to do the math is more accurate.

using tcl expr

[expr (@Value(E_x)-@Value(S_x))/sqrt(pow(@Value(E_x)-@Value(S_x),2))]

yields the correct -1

I always thought fme waas using the same math engine, tcl one, but appearently not?

so

This post is closed to further activity.
It may be an old question, an answered question, an implemented idea, or a notification-only post.
Please check post dates before relying on any information in a question or answer.
For follow-up or related questions, please post a new question or idea.
If there is a genuine update to be made, please contact us and request that the post is reopened.

2 replies

M
Forum|alt.badge.img+5

OK, the easy part first is that you need to set View > Display Full Precision in the Data Inspector. Then it will show the true value and not round it off to -1

As for the other, I found that if I create:

result1 = (@Value(E_x)-@Value(S_x))

result2 = @sqrt(@pow(@Value(E_x)-@Value(S_x),2))

result3 = result1/result2

Then result3 is -1 (whereas if I put the whole thing in one expression as you do, then I get -0.99999999999999967).

So something is rounding the data. As far as I know, yes FME does use the Tcl math engine. So I will query both points with our developers and see what they say.

M
Forum|alt.badge.img+5

If you're still following this thread - can I ask:

- what platform you are on?

- what build and version of FME are you using (32/64-bit?)

- what tcl version did you use? 8.4, or 8.5 and later?

I have a feeling this is a complex scenario. If you notice, my result is different to your result, suggesting it is platform dependent!

Community Stats

32,208
Posts
122,306
Replies
39,892
Members

Latest FME

Terms & ConditionsCookie settingsAccessibility statement
A product of
Safe Software respectfully acknowledges that we live, learn and work on the traditional and unceded territories of the Kwantlen, Katzie, and Semiahmoo First Nations.