Commit Graph

728 Commits

Author SHA1 Message Date
Roger Labbe
afe168c7b3 Removed warning that chapter cannot be read.
Code is working well now.
2015-06-30 20:16:24 -07:00
Roger Labbe
24e70d7c77 Added MM filter bank. 2015-06-30 20:14:39 -07:00
Roger Labbe
1b92397175 Checkpoint. Added a lot to the particle filter chapter. 2015-06-28 13:38:12 -07:00
Roger Labbe
7cf3980e8c Made covarinace ellipse transparent.
The solid fill made the plots harder to read, so I set them
to no fill color.
2015-06-28 11:20:27 -07:00
Roger Labbe
966c6028c3 Removed erroneous chart from PF chapter.
There was a really misleading chart in the old pf material.
It was destined to be deleted, but I didn't want peoople to
be mislead by it if browsing the chapter on github.
2015-06-27 08:41:06 -07:00
Roger Labbe
fbcef07499 Adding Particle filter chapter.
This is an early, incomplete draft, but it is time to get it into
source control so I can track changes. Not ready for public
consumption.
2015-06-27 08:37:14 -07:00
Roger Labbe
960771b4c4 Added algorithm pseudocode to help link chapters.
I don't think the commonality of the algorithms was made as
expllicit as it could be. More work needs to be done in this regard.
2015-06-25 18:39:57 -07:00
Roger Labbe
10a62649df Added chart of multiple Gaussians 2015-06-24 21:00:57 -07:00
Roger Labbe
46b3be3139 Fixed missing import. 2015-06-23 21:30:23 -07:00
Roger Labbe
0a33395005 Fixed variance units, and missing import
I still had several places where i used units of meters for the
variance, when of course it should be squared.
2015-06-22 21:19:23 -07:00
Roger Labbe
5f593639d8 Altered plot styles.
Lines seemed to thick to see what is happening when they are close
to each other. Changed line width to 2.
2015-06-22 17:56:14 -07:00
Roger Labbe
51c9a8283e Wholesale changes to connect chapters together.
I made a lot of changes so that each chapter makes clear that
they are all implementing the same basic bayesian algorithm.

This required a lot of editting, and it doesn't make sense to
try to do that atomically, hence this huge check in.

I made a lot of edits, and haven't copy editted anything. i'm
sure I introduced a lot of problems and discontinuities.
2015-06-20 15:52:16 -07:00
Roger Labbe
5ffe5c67eb Added UKF/EKF comparison.
Had to adjust some examples and rerun notebooks in several places
to account for some code chantes that I made.
2015-06-17 09:12:31 -07:00
Roger Labbe
87e6f5c6e7 Removed testing code. 2015-06-17 08:04:06 -07:00
Roger Labbe
adba3e5915 Renamed a function. 2015-06-16 22:10:42 -07:00
Roger Labbe
41673aab60 Added reference to computing PDF notebook. 2015-06-16 22:07:09 -07:00
Roger Labbe
d280a9f344 Added supporting notebooks subfolder.
Wrote a notebook explaining how to plot PDFs and run monte
carlo simulations and compute their results.
2015-06-16 21:56:21 -07:00
Roger Labbe
1633d9cda9 Added fixed lag section. 2015-06-13 20:05:10 -07:00
Roger Labbe
d574383175 Updated required version for FilterPy. 2015-06-13 13:23:06 -07:00
Roger Labbe
4adfc8ca50 Changes in test code for ukf - not for book. 2015-06-13 13:22:43 -07:00
Roger Labbe
1152bf053d Fixed angle normalization bug. 2015-06-13 13:21:17 -07:00
Roger Labbe
4c712918b8 Updated for UKF bug fixes. Added summary. 2015-06-13 13:20:29 -07:00
Roger Labbe
61d6931b3a Before major UKF change.
this is not an atomic check in. Probably won't even run.Just saving
my place before making some big changes.
2015-06-13 09:23:11 -07:00
Roger Labbe
aadc5abaf6 Added robot steering example to UKF. 2015-06-09 19:39:39 -07:00
Roger Labbe
3c3a5e5bf6 Final updates before merge to trunk.
Updated to work with FilterPy 0.0.19.
2015-06-08 19:21:18 -07:00
Roger Labbe
88d200b9a8 Robot localization example added.
Still needs work; code is messy, and Q is not computed. But I am
getting correct results.

Also need to expand explanations and scenarios.

Need to remove now redundant explanations in EKF chapter, since
they exist in the UKF chapter.
2015-06-08 07:23:52 -07:00
Roger Labbe
f8057cd917 UKF chapter updated for Merwe parametization.
Chapter is not complete; checking in because I am beginning to alter
the formulation of unscented_transform in filterpy.
2015-06-07 17:03:35 -07:00
Roger Labbe
9c09133b5b Working with Merwe formulation. 2015-06-07 16:43:38 -07:00
Roger Labbe
758bd2c7f4 Working for UKF rename.
Most of the text is wrong, but changed code to use the
renamed ScaledUnscentedKalmanFilter.

Checking in with bad text because I am in the process of changing
FilterPy to use a class for the sigma points to make it easier
to change the sigma point generation, leaving us with one
UKF class instead of several.
2015-06-06 19:49:56 -07:00
Roger Labbe
85eccf7a21 Moved extraneous chapters to appendix. 2015-05-31 15:20:44 -07:00
Roger Labbe
2c8a2d0a8e Updated robot localization example.
Put into a function, added parameters so we can run the alg
with different landmarks and sigmas.
2015-05-31 14:30:53 -07:00
Roger Labbe
54bce9d7a0 Added EKF robot localization example.
Still needs a lot of explanation; mostly the implementation is there
for now.
2015-05-30 14:44:49 -07:00
Roger Labbe
36d45b35bb latex needed a space after \pm for some reason.
This rendered fine in the notebook, but pdf generation
choked.
2015-05-28 17:22:17 -07:00
Roger Labbe
2271c96dbd Merge branch 'master' of https://github.com/rlabbe/Kalman-and-Bayesian-Filters-in-Python 2015-05-24 17:03:13 -07:00
Roger Labbe
449579fd64 Added PDF to ignore
We don't check in the PDF anymore because it makes the repository
far too big.
2015-05-24 17:02:30 -07:00
Roger Labbe
163052bcb1 Minor edits.
changed variable name, cleaned up a few expressions.
2015-05-24 17:00:26 -07:00
Roger Labbe
a30013c00d Expanded description of multiple updates.
Put in context of future EKF localization problem where we
perform several updates in a row, one for each landmark.
2015-05-24 16:55:28 -07:00
Roger Labbe
78d5644769 Merge pull request #44 from dduong42/master
Add `dt` to the argument list.
2015-05-23 12:16:49 -07:00
Daniel DUONG
4c05ec2b49 Add dt to the argument list. 2015-05-23 21:12:16 +02:00
Roger Labbe
6334c3eaa9 Merge pull request #43 from dduong42/patch-1
Fix installation link in README
2015-05-23 10:59:51 -07:00
Daniel Duong
845511b9d6 Fix installation link in README 2015-05-23 19:54:24 +02:00
Roger Labbe
3796453e1c Switched train speed to value evenly divisible.
This makes the example clearer since there is no round off error.
2015-05-23 09:43:13 -07:00
Roger Labbe
3cce5adf64 Fixed math for train. 2015-05-23 09:27:56 -07:00
Roger Labbe
4f1558b12a Added link to PDF.
I removed the PDF from the repository, but people still want the PDF. I put a copy on my google drive, and another on PDF archive. Not sure which, if either, I will maintain for the long term.
2015-05-22 12:44:05 -07:00
Roger Labbe
e18106920b Adjusted latex font.
The Jacobian matrices were rendering way to small to read, so I
changed the font and increased the font size.
2015-05-21 15:47:46 -07:00
Roger Labbe
7725ab03f5 Improved nbviewer documentation. 2015-05-19 20:15:18 -07:00
Roger Labbe
26fe2968e4 Removed links to PDF.
I removed the PDF from the repository because git does poorly with
binary objects.
2015-05-19 20:12:32 -07:00
Roger Labbe
8a9ecc026e Revisions based on Radford's comments. 2015-05-19 18:25:54 -07:00
Roger Labbe
2e0e513912 Merge pull request #35 from pquentin/master
Fix issue #16 properly without centered headers
2015-05-19 18:00:14 -07:00
Roger Labbe
8399c6d4d6 Merge pull request #39 from pkulchenko/extended-kalman-chapter-typo
Fixed small typo in Extended Kalman chapter.
2015-05-19 17:59:47 -07:00