[Kst] How to compute a 2-sigma scalar?

Brisset, Nicolas Nicolas.Brisset at eurocopter.com
Wed Jul 25 10:25:07 UTC 2012


we are using Kst to analyse some data, where we basically do the following for each value:

-          Compute the error and absolute error against a known reference

-          Compute the histogram (in percent) of the absolute error

-          Integrate the histogram with the cumulative sum plugin and a step of 1

-          Plot the cumulative sum against the bin vector of the histogram: you get a nice curve showing how many percent of the cases are below a given error value

-          We compute an equation [Cumulated error distribution]>95 and use it as X-marker in the plot => we get a nice visual feedback of the 2-sigma error of that value

Now the problem is, we have a whole bunch of curves spread over many tabs. So I was thinking we could add a tab with text labels pulling in some interesting scalars, so that we have a nice overview, typically the 2-sigma error of each value. But I can't find a way to get to the interesting scalar. The MinPos of the >95 curve is not useful because it does not hold the last pos of the min, but the first one. MaxPos could do it, but it doesn't exist...
It would be possible to write a plugin for that, but that somehow sounds like overkill.
Does anybody have a good idea?


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kst/attachments/20120725/fd3e037a/attachment-0001.html>

More information about the Kst mailing list