Kalman filter code?

[Mike Brenner]

Steve and Bjorn, 

Kalman filters are interesting. However, a Kalman filter is not a single
algorithm. It depends on the model of the system being built. However
the idea of feeding back the variance matrix, etc., is common to all the
algorithms.

Python might not be the right language to do this in realtime because it
involves large numbers of matrix operations repeated frequently.
However, I believe Python is probably the right language to DEVELOP it
in, or to SPECIFY the algorithm in.


> I am new to Python.

Python is easy to learn and the arithmetic and arrays are particularly
easy. Just have to use the right editor and never cut and paste,
otherwise you lose all your indenting.


> I know nothing about Kalman filtering.

Yes, you will have to learn how to set up a feedback model and put the
numbers into the matrices that come from the model, but it is learnable.


> ... you are talking either constant-system discrete/continuous time filter discrete time (time varying is harder) ...

Perhaps a really good reason to do it in Python is the easy with which
dynamic models could be modeled. In other words, we could take a box
diagrammer and GENERATE kalman filters for any system we choose. 

While there are expensive products on the market that do some of this,
it would be cool to have a fully changeable model, for additional
reasons beyond Kalman filters. 

For example, the same graphical inputs that create the feedback lines
between the boxes for the Kalman filter generator could generate things
like finite state machines, transducers, linear arrays of Kalman filters
called Kaman filters (without the L), IIF filters, colored Petri Nets,
BNF grammars, dataflow processors, and eCommerce middle tiers to process
transactions.

Mike Brenner