Repository

## rlabbe/Kalman-and-Bayesian-Filters-in-Python

Kalman Filter book using Jupyter Notebook. Focuses on building intuition and experience, not formal proofs. Includes Kalman filters,extended Kalman filters, unscented Kalman filters, particle filters, and more. All exercises include solutions.
13127 3373 463 270
Issue

## SIR Particle Filter time series

• 000
I wanted to ask how it is possible to add the Air Passenger data set to the SIR filter example and make a prediction for the future and plot all the future steps. Is it something like adding (-4) to the predict 4 steps and add the data to the landmarks np.array? Could you add/show this to the notebo...
Issue

## Chapter 6: making Q very large

• 000
In the last paragraph it says "if you make it (Q) too large the filter will fail to respond quickly to changes". This is wrong, if the process covariance is large then we will listen more to the measurements making it quicker to changes.
Issue

## Chapter 3: Variance of a Random Variable

• 000
The derivation of variance of a random variable in chapter 3 could be clarified. At the moment, it is stated that the equation for computing variance is: $$\mathit{VAR}(X) = \mathbb E[(X - \mu)^2]$$ And that the formula for expected value is $\mathbb E[X] = \sum\limits_{i=1}^n p_ix_i$, which can be...
Issue

## Chapter 3: It should say posterior not prios

• 000
In Chapter 3 in the last sentence of the first paragraph of the section Total Probability Theorem i think it should say posterior instead of prior. The probability of being at any position ๐ at time ๐ก can be written as ๐(๐๐ก๐) . We computed that as the sum of the prior at time ๐กโ1 ๐(๐๐กโ1๐)...
Issue

## Incorrect convolution math formula in chapter on Discrete Bayes

• 000
https://nbviewer.org/github/rlabbe/Kalman-and-Bayesian-Filters-in-Python/blob/master/02-Discrete-Bayes.ipynb In the Discrete Bayes chapter, we have the formula for the discrete convolution: For t=0 this always yields just one value that is being added. This seems incorrect to me. Shouldn't the corr...
Issue

## Question: what is the recommended way to introduce seasonality to Kalman filter?

• 000
Can an Unscented Kalman Filter encode seasonality in the data? If so, what is the recommended way of doing it? Are there any references out there for introducing seasonality on Kalman Filter?
Issue

## discrete_bayes_sim has unused paramenter prior

• 000
In 02-Discrete-Bayes.ipynb in the chapter The Discrete Bayes Algorithm you define the following function with the signature discrete_bayes_sim(prior, kernel, measurements, z_prob, hallway). The parameter prior is unused.
Issue

## Non normal noise

• 000
First thank you for your amazing book. I went through most of it and it had been of invaluable help. I was wondering if you had some advice when dealing with non normal process or sensor noise. You always model noise as discrete white noise. How would you adapt the UKF with a noise that is following...
Issue

## Get velocity and acceleration from noisy 2D position

• 000
Hi, this is an amazing content. I would like to filter my 2D signal to obtain both the velocity and acceleration at a time t, in order to estimate a CTRA model. Which method would you recommend to me? I am using the Argoverse Motion Forecasting dataset. An example of what I am trying to do is as fol...
Issue

## Chapter 03 - central limit theorem

• 000
First of all, thank you very much for this learning material, I really wish there were similar tutorials for other engineering areas! I think the central limit theorem is slightly off in chapter 3: What does this curve mean? Assume we have a thermometer which reads 22ยฐC. No thermometer is perfectly...