Systematic Error due to Accuracy of Number System?

Catkun · March 13, 2015, 05:21

Dear all,

I am writing some code in matlab that uses a numerical method called WENO to solve Euler equation. However, there is always a error of size around 0.01 no matter how I changed the parameters. Then I wrote some tests and noticed that in one place Matlab does not return a right number: it is supposed to return -1e-14 or so but it returned +1e-15. It seems that this error is systematic and lead to the final 0.01 error in result.

I wonder if anyone encounter similar situations or that it's just my code's problem. Thanks for any thoughts.

fluid23 · March 13, 2015, 12:38

1e-15 is on the order of machine epsilon even if you are using double precision.

I would think the Matlab user forum would be a better place to ask this.

http://www.mathworks.com/matlabcentral/answers/

March 13, 2015, 05:21	Systematic Error due to Accuracy of Number System?	#1
Catkun New Member Cat Join Date: Aug 2014 Posts: 6 Rep Power: 11	Dear all, I am writing some code in matlab that uses a numerical method called WENO to solve Euler equation. However, there is always a error of size around 0.01 no matter how I changed the parameters. Then I wrote some tests and noticed that in one place Matlab does not return a right number: it is supposed to return -1e-14 or so but it returned +1e-15. It seems that this error is systematic and lead to the final 0.01 error in result. I wonder if anyone encounter similar situations or that it's just my code's problem. Thanks for any thoughts.

March 13, 2015, 12:38		#2
fluid23 Senior Member Matt Join Date: Aug 2014 Posts: 947 Rep Power: 17	1e-15 is on the order of machine epsilon even if you are using double precision. I would think the Matlab user forum would be a better place to ask this. http://www.mathworks.com/matlabcentral/answers/ Last edited by fluid23; March 13, 2015 at 12:39. Reason: spelling correction

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
ERROR #001100279 has occurred in subroutine ErrAction.	smnaryal	CFX	11	December 20, 2017 16:32
same geometry,structured and unstructured mesh,different behaviour.	sharonyue	OpenFOAM Running, Solving & CFD	13	January 2, 2013 22:40
CFX11 + Fortran compiler ?	Mohan	CFX	20	March 30, 2011 18:56
[snappyHexMesh] snappyHexMesh won't work - zeros everywhere!	sc298	OpenFOAM Meshing & Mesh Conversion	2	March 27, 2011 21:11
Unaligned accesses on IA64	andre	OpenFOAM	5	June 23, 2008 10:37