Kalman Innovation Value

Problem

Compute the innovation, or measurement residual, in a Kalman filter: v = obs - h * prior, where prior is the predicted state and h is the observation matrix.

The innovation is the new information an observation carries and drives the correction applied to the state estimate.

Hint

Compare the observation with the value the model predicted for it.

Test cases

case 1

in: [10,1,8]

expected: 2

case 2

in: [5,2,3]

expected: -1

case 3

in: [0,1,0]

expected: 0

case 4

in: [7,0.5,10]

expected: 2

Starter codePython

def kalman_innovation(obs, h, prior):
    # Your code here
    pass

4 test cases · run your code against them in the editor