Kalman-0.1.0.1: src/Kalman.hs
{-# OPTIONS_GHC -Wall #-}
{-# OPTIONS_GHC -fno-warn-name-shadowing #-}
{-# OPTIONS_GHC -fno-warn-type-defaults #-}
{-# OPTIONS_GHC -fno-warn-unused-do-bind #-}
{-# OPTIONS_GHC -fno-warn-missing-methods #-}
{-# OPTIONS_GHC -fno-warn-orphans #-}
{-# LANGUAGE DataKinds #-}
{-# LANGUAGE ScopedTypeVariables #-}
{-# LANGUAGE TypeOperators #-}
{-# LANGUAGE TypeFamilies #-}
module Kalman (
extKalman
)where
import GHC.TypeLits
import Numeric.LinearAlgebra.Static hiding ( create )
import Data.Maybe ( fromJust )
extKalman :: forall m n .
(KnownNat m, KnownNat n, (1 <=? n) ~ 'True, (1 <=? m) ~ 'True) =>
R n -- ^ Prior mean
-> Sq n -- ^ Prior variance
-> L m n -- ^ Observation map (represented as a matrix)
-> Sq m -- ^ Observation noise
-> (R n -> R n) -- ^ State update function
-> (R n -> Sq n) -- ^ A function which returns the
-- Jacobian of the state update
-- function at a given point
-> Sq n -- ^ State noise
-> [R m] -- ^ List of observations
-> [(R n, Sq n)] -- ^ List of posterior means and variances
extKalman muPrior sigmaPrior bigH bigSigmaY
littleA bigABuilder bigSigmaX ys = result
where
result = scanl update (muPrior, sigmaPrior) ys
update :: (R n, Sq n) -> R m -> (R n, Sq n)
update (xHatFlat, bigSigmaHatFlat) y =
(xHatFlatNew, bigSigmaHatFlatNew)
where
v :: R m
v = y - (bigH #> xHatFlat)
bigS :: Sq m
bigS = bigH <> bigSigmaHatFlat <> (tr bigH) + bigSigmaY
bigK :: L n m
bigK = bigSigmaHatFlat <> (tr bigH) <> (inv bigS)
xHat :: R n
xHat = xHatFlat + bigK #> v
bigSigmaHat :: Sq n
bigSigmaHat = bigSigmaHatFlat - bigK <> bigS <> (tr bigK)
bigA :: Sq n
bigA = bigABuilder xHat
xHatFlatNew :: R n
xHatFlatNew = littleA xHat
bigSigmaHatFlatNew :: Sq n
bigSigmaHatFlatNew = bigA <> bigSigmaHat <> (tr bigA) + bigSigmaX
inv :: (KnownNat n, (1 <=? n) ~ 'True) => Sq n -> Sq n
inv m = fromJust $ linSolve m eye