KLE expansion for the generation of stochastic processes
Hi,
how about in addition to the spectralrepresentation we implement the Karhunen-Loève Expansion for stationary and non-stationary stochastic signal generation? This method should then be related to the abstract model StochasticProcessModel.
Actually, it would be great if some of us who are interested and comfortable with coding could start thinking about how we might conceptually and prospectively integrate random fields into the framework. After all, these tools (spectralrepresentation and prospectively the KLE expansion) aren’t just meant for stochastic processes — they’re also highly relevant for modeling random fields.
Greetings.
Yes to KLE and random fields!
I was involved with KLE within OpenCossan. Some stuff in there was debatable but I can explain to you what we did. I wouldn't convert that code into here though.