[Kst] Event detection in kst ?

Nicolas.Brisset at eurocopter.com Nicolas.Brisset at eurocopter.com
Wed Apr 7 16:05:45 CEST 2004


I've built kst-0.94 recently and played with it a bit. It looks really good 
and I'm hoping it can soon replace xmgrace, whose interface is really clumsy 
and not so usable even though the feature set is nice :-) I'm currently 
building kst 0.97 to see how it has changed...

Speaking about all this, one feature I often miss (in xmgrace and other 
plotting programs) is a "event detection" feature. When you have huge 
datasets (like hundreds of variables and thousands of samples for each of 
them), it is not feasible to look at everything. What we want to do is have 
an "event scanner" look for specific events in the data, and make a small 
report containing among others the time at which an event occured. We can 
then use this information to investigate more closely what happens around 
that time.
It is a kind of datamining operation, where events would be "raising edge on 
boolean Bool1" or "Var1 and Var2 differ by at least 1.5", etc. You get the 
I was wondering whether this is planned (or already implemented :-)) ?



More information about the Kst mailing list