WCA / News: Recent posts

All statistical analysis exploit the new initial setting statement.

The goal of this statement is to simulate the tuning of a performance by choosing a part out of a list of available parts. Obviously, this tuning only affects the initial dispersion of the performance.

The statement
set f to 0.5 with c in {0.829:0.867:0.905:0.943:0.981:01.019:1.057:1.095:1.133:1.171}
will implies that the function f will be adjusted as close as possible to 0.5 by using for the variable c one of the value between {}.
This operation only affects the initial drifts of the function f.... read more

In the WCA project, you will now find an extension for MathCAD.
It extends MathCAD so that it becomes possible to performed easily a Monte-Carlo worst case analysis. The drifts are loaded from a database so that all your WCA will use the same drifts.

There is now a wiki for the WCA project

The function if(z1,z2,z3) is added.
It returns z2 if the real part of z1 is strictly positive, it returns z3 otherwise.

Remove some incompabilities with ansi C99.

Some bug fixes.

I notice that the old archive can't be uncompress. So I just uploaded a new one.

Function without parameter can now be referenced without the () suffix.

I just release the version 0.2 of WCA.
The differences with the version 0.1 are
-Computes sampling parameters for the statiscal worst case.
-New command line option -c to define the desired confidence in the statistical worstcase.
-Sampling parameter are printed in the report header.
-Plots are computed on 100 points instead of 20.
-Worst-case analysis are computed on at Tnom in addition to Tmin and Tmax.... read more

The version 0.1 sources arrived!