5857635ca1
Fixed renaming error with NonlinearFactors sometimes becoming NonGaussianFactors
2009-11-12 16:47:12 +00:00
c7b86cec97
Renamed ConditionalGaussian -> GaussianConditional
2009-11-12 16:41:18 +00:00
77a1754b69
Renamed LinearFactor -> GaussianFactor, LinearFactorGraph -> GaussianFactorGraph
2009-11-12 16:16:32 +00:00
1ae6bb4030
Added push_fron convenience method to add ConditionalGaussians into a Bayes net with much less clutter. Modernized some very old tests in the process.
2009-11-12 06:09:03 +00:00
f677341108
Moved inference methods to new compilation unit. Added [factor], and [marginalize] now returns a factor graph.
2009-11-12 04:56:30 +00:00
c85f5445db
Re-organized makefile
2009-11-12 04:54:31 +00:00
81a353dd2c
eliminate is now const
2009-11-12 04:53:28 +00:00
a38ebece1e
New compilation unit that collects generic, templated inference methods that go between factor graphs and Bayes nets. These used to be in BayesNet-inl.h and FactorGraph-inl.h
2009-11-12 04:52:40 +00:00
4a7e05fffa
change all print() to print(string s) and create SymbolicFactor
2009-11-11 22:45:08 +00:00
70efccefbc
VSLAMFactor Testable
2009-11-11 21:09:43 +00:00
8c695a778c
Implemented a rank deficient constraint test in testLinearFactor: note that it is possible to reorder elimination to make a system solvable or rank-deficient
2009-11-11 19:00:39 +00:00
ab65dfb4dd
Adding include file stdio.h to resolve compiling issue when upgrading to Ubuntu 9.10. The printf() compiles on ubuntu 9.04 and mac machines.
2009-11-11 17:15:17 +00:00
f51614813e
Improved weighted eliminate to handle arbitrary linear equality constraints
2009-11-11 14:42:09 +00:00
d585a329dc
Added MATLAB-friendly versions of eliminate and optimize in LinearFactorGraph (with trailing underscore). Also fixed some MATLAB code.
2009-11-11 07:14:13 +00:00
ac840d6f0a
1) DCVectorConfig I believe was obsolete as well as any conversion from and to Vectors, which I removed
...
2) load and flush are gone.
3) I made a VSLAMConfig "Testable".
4) I added an explicit template instantiation to avoid having to include -inl.h files
5) I added some more test code in testVSLAMFactor
2009-11-11 05:14:03 +00:00
00ff066b33
Fixed compile error
2009-11-11 05:13:25 +00:00
8d2d48d252
made removeAndCombineFactors a function, not a method
2009-11-11 05:12:45 +00:00
710d396d0b
cleaned up a few includes to make VSLAMGraph work properly after the move into gtsam.
2009-11-10 22:49:22 +00:00
dc3b85e43d
Todo: make wrapper for GaussianBayesNet. For now eliminate returns void in gtsam.h
2009-11-10 21:15:28 +00:00
08c9718b12
Moving VSLAMConfig into gtsam. VSLAMFactor lived there already, but was using VectorConfig directly... now using VSLAMConfig.
2009-11-10 20:19:00 +00:00
0d928035e0
added empty constructors for serialization, also made sure SymbolicConditional.h was installed
2009-11-10 14:55:07 +00:00
1f15650da0
Added serialization of base class
2009-11-10 14:44:53 +00:00
ddc0173671
Basic handling of constraints now works in factor graphs, assuming there is only one constraint on any given variable.
2009-11-10 04:36:07 +00:00
a7b711db37
Removed constraint variations on graphs/factors
2009-11-10 04:33:39 +00:00
2c37c94b5d
Replaced the householder transform with the weighted system
...
Removed constrained components from makefile, they will disappear shortly
2009-11-09 21:34:20 +00:00
03985d16f6
Fixed equality checking with NaNs in Matrix
...
removed extraneous print statement in Vector
2009-11-09 16:57:30 +00:00
01ee9246d7
fix Matlab toolbox to work with Frank's recent changes to gtsam.
2009-11-09 16:53:13 +00:00
1ae81b59d8
Fixed vector comparisons that will erroneously return true when there are NaN values.
2009-11-09 16:47:37 +00:00
a3de1964d7
BIG CHANGE:
...
1) eliminate methods no longer return a shared pointer. Shared pointers are good for Factors and Conditionals (which are also non-copyable), because these are often passed around under the hood. However, a BayesNet is simple a list of shared pointers and hence does not cost a lot to return as an object (which is compiler-optimized anyway: there is no copy). So, the signature of all eliminate methods changed to simply return a BayesNet<> object (not a shared pointer).
2) GaussianBayesNet::optimize is now replaced by optimize(GaussianBayesNet) and returns a VectorConfig and not a shared pointer
3) GaussianBayesNet and SymbolicBayesNet are now simply typedefs, not derived classes. This is desirable because the BayesTree class uses templated methods that return BayesNet<Conditional>, not a specific BayesNet derived class.
2009-11-09 07:04:26 +00:00
7bd40e836d
Cleaned up and added test for VSLAMFactor
2009-11-09 04:46:34 +00:00
e340178de5
Easy constructor
2009-11-09 04:46:08 +00:00
dd1b023ca9
Fixed special case for root
2009-11-09 04:45:38 +00:00
cc22e82ca6
Joint densities (covariances) now done. Was exceedingly easy :-)
2009-11-09 00:13:44 +00:00
3464c6a36f
Create from a single string
2009-11-08 22:51:29 +00:00
86173b66af
Clique marginal and dramatically simplified single variable marginal.
2009-11-08 22:51:12 +00:00
10e618f360
New function marginals. Also: combine entire Bayes nets with push_back and push_front. And finally: some convenience constructors in GaussianBayesNet.
2009-11-08 22:50:26 +00:00
58007a8167
Working recursive (but not cached) shortcut.
2009-11-08 04:41:01 +00:00
d1b8925617
Construct symbolic version of any Bayes net
2009-11-08 04:39:54 +00:00
532be063f9
new templated function combine
2009-11-08 02:50:19 +00:00
93da9e0f2c
pop_front
2009-11-08 02:49:18 +00:00
cc5a2c3183
Renamed double-templated functions to _eliminate and _eliminateOne, and created FactorGraph-specific eliminateOne methods to make life easier
2009-11-07 21:03:30 +00:00
df3e5f2416
BIG: eliminate and eliminateOne now doubly templated functions, not methods.
...
Minor: Standardized on new shared_ptr naming convention:
shared_factor -> sharedFactor
conditional_ptr -> sharedConditional
node_ptr -> sharedClique
2009-11-07 19:31:39 +00:00
f5d186522e
Standardized on new shared_ptr naming convention
2009-11-07 17:24:05 +00:00
6339b5d19d
Made Node public
2009-11-07 15:58:45 +00:00
7ef9ed950b
Improved performance of matrix scaling in LinearFactor::matrix_augmented
...
Added matrix scaling functions to scale either the rows or columns by values from a vector
2009-11-06 13:43:39 +00:00
cd313e2f82
operator[] to retrieve clique
2009-11-06 05:55:29 +00:00
e1716a39cd
Two changes: LinearFactor::sparse and LinearFactorGraph:sparse, and renamed VariableSet -> Dimensions, which is now a map from keys to integer variable dimensions. Merged in from the "sparse" branch created with Viorela.
2009-11-06 05:43:03 +00:00
fb74ef03b2
Added to LinearFactor a matrix_augmented() function to get a single augmented matrix [A b]
...
Removed extra copying in LinearFactor::eliminate()
2009-11-05 15:08:58 +00:00
6c5603de0c
Fixed remaining LinearFactor unit tests
...
Added equality check for Variable
2009-11-05 13:52:12 +00:00
d9289d14b3
marginals on any scalar now work
2009-11-05 08:06:32 +00:00
beabb62f29
More informative print
2009-11-05 08:05:54 +00:00
bd54c39a73
Fixed bug in smoother example
2009-11-05 08:05:34 +00:00
e87c19ed7a
ConditionalGaussian now stores sigmas
2009-11-05 06:59:59 +00:00
88b8f4e16b
fixed unit test
2009-11-05 06:34:43 +00:00
cabcda5a96
marginal compiles and runs for frontal node in clique
2009-11-05 06:30:50 +00:00
7f516394df
Added dummy argument to work around compile conundrum
2009-11-05 06:30:10 +00:00
06f07420d1
back()
2009-11-05 06:29:32 +00:00
11fcd5a69d
KISS: just a map of pointers in BayesTree now
2009-11-05 05:29:47 +00:00
3d334401f5
Constructor from BayesNet now in FactorGraph base class
...
FactorGraph::eliminate
Removed eliminate_partially
2009-11-05 04:56:59 +00:00
e650676da7
Moved header to .h to fix MATLAB toolbox compilation
2009-11-05 03:51:08 +00:00
3b258be8c4
about a dozen changes to make Matlab toolbox work again, but one error remains having to do with the forward declaration of class GaussianBayesNet.
2009-11-05 03:26:04 +00:00
0ef7fe4a01
changed case of foreach ( it was capitals before and will not compile on linux )
2009-11-04 22:19:16 +00:00
e2414561be
Merged r895:900 from branch weightedQR - LinearFactorGraph now has sigmas and ConditionalGaussian now has precisions
2009-11-04 20:59:16 +00:00
4865edb883
Removed debug code, added marginal function
2009-11-04 03:22:29 +00:00
e9d942f81e
BayesNet is now list-based for fast bi-directional access
...
SLOW O(n) random access operator[key] provided
(should maybe be called [at] as it does bounds checking)
I also fixed a bug in equals.
2009-11-03 06:29:56 +00:00
eab038651e
Renamed BayesNet::insert -> push_back. BayesTree now uses Bayes nets as nodes.
2009-11-02 05:17:44 +00:00
a8d267c4ca
Small change necessitating lots of edits: Conditionals now include key of random variable
...
This simplifies Bayes nets quite a bit. Also created a Conditional base class, derived classes ConditionalGaussian and SymbolicConditional
Finally, some changes were needed because I moved some headers to .cpp
2009-11-02 03:50:30 +00:00
943b692a6b
BIG CHANGE: I got rid of the BayesChain/ChordalBayesNet classes and we now simply have a BayesNet class. It will just happen to be chordal when it is the result of an elimination. This will simplify a lot of things.
...
The main renaming that happened is
BayesChain -> BayesNet
ChordalBayesNet -> GaussianBayesNet == BayesNet<ConditionalGaussian>
SymbolicBayesChain -> SymbolicBayesNet == BayesNet<SymbolicConditional>
2009-10-31 19:53:20 +00:00
921cb0a8fc
Got rid of (defunct) "dump" methods
2009-10-31 17:13:36 +00:00
c046fed37c
Bayes tree constructor implemented and tested with ASIA, as well as smoother example from frankcvs meeting
2009-10-31 16:57:36 +00:00
1e5a2d692a
const
2009-10-31 16:56:06 +00:00
040f407b0b
Removed constructor, use eliminate
2009-10-31 16:55:44 +00:00
0f688302ad
Fixed bug in [parents]
2009-10-31 16:55:17 +00:00
68d2f81f0a
Smoother now creates x1...xT, not x0 anymore
2009-10-31 16:54:38 +00:00
fd5f43092f
Ordering is now a list and Testable
2009-10-31 15:24:22 +00:00
e1f14b34c3
Implemented [equals] using binary predicate and STL's equal
2009-10-31 14:12:41 +00:00
b9ceca7dc6
Templated binary predicate [equals] for use with STL's equal(...)
2009-10-31 14:11:48 +00:00
53890c4ba6
Symbolic Bayes Tree successfully constructed
2009-10-31 05:12:39 +00:00
cefeca149b
Moved template definitions to BayesTree-inl.h
2009-10-30 13:03:38 +00:00
0d7d0e2161
Explicit template instantiation in .cpp files, so we can link with the code without having to include -inl.h everywhere...
2009-10-30 05:45:22 +00:00
83e5286710
ChordalBayesNet is now derived from BayesChain
...
Explicit template instantiations
2009-10-30 04:54:11 +00:00
dd3795ad5a
Symbolic eliminate now works, new compilation unit SymbolicFactor
2009-10-30 03:48:32 +00:00
65b949c008
Fixed flaw in pseudoinverse calculation, and updated tests for Matrix and Vector
2009-10-29 20:12:07 +00:00
80b162a412
LinearFactorGraph::eliminate_one is now FactorGraph::eliminateOne<ConditionalGaussian>
...
Symbolic version FactorGraph::eliminateOne<SymbolicConditional> also implemented and tested
2009-10-29 14:34:34 +00:00
4c48bb08e1
Vector and Matrix updated with improved weighted householder operations and more tests.
2009-10-29 12:52:27 +00:00
a513ae0287
Changed names and moved two LinearFactorGraph functions to FactorGraph
...
Added SymbolicFactorGraph compiulation unit and unit tests
Added symbolic combine constructor
2009-10-29 05:39:13 +00:00
f0c23a2828
Moved [factors] to FactorGraph
2009-10-29 04:49:13 +00:00
b6cee73571
Significant change: Made FactorGraph templated on Factor only, and moved error and probPrime to derived classes
...
Moved find_and_remove_factors to base class
Added and tested symbolic factor graph constructor and conversion from any factor graph type
2009-10-29 04:11:23 +00:00
1f792a53ea
Made some progress on symbolic analysis
2009-10-28 02:57:38 +00:00
50763ef764
Prototype for Bayes nets for Manohar
2009-10-27 20:44:40 +00:00
bb15697cf9
Prototype framework for Bayes chain (new name for Chordal Bayes net) and Bayes tree class hierarchy
2009-10-27 20:23:19 +00:00
94da98411c
Fixed NULL problem
2009-10-27 20:22:02 +00:00
e37e771c8e
My last commit broke the linearFactorN unit test, fixed now.
2009-10-27 15:31:45 +00:00
cb22a0bc52
Renamed variables in linearFactorN unit test
2009-10-27 15:13:13 +00:00
37bc303492
Added weighted Householder transforms that use precisions perform QDR factorization. Functions create a weighted vector pseudoinverse, and then use the pseudoinverse to substitute a solution into system.
2009-10-27 14:21:22 +00:00
626d06905c
Removed obsolete optimize call and documented better
2009-10-27 14:14:36 +00:00
4d9ff77249
moved timing example here from EasySLAM
2009-10-27 13:34:36 +00:00
422169873c
moved [symbol] function to LinearFactor
2009-10-27 13:33:44 +00:00
11f0d04cb6
made all classes that currently have an assert_equal "Testable"
...
- derive from testable as in class Point2 : Testable<Point2>
- moved print and equal declarations in .h right after the constructor
- similarly, moved implementations after constructors in .cpp file
- removed obsolete assert_equal
2009-10-26 19:26:51 +00:00
3b30fe50b0
Moved print and equals to -inl.h
2009-10-25 22:27:18 +00:00
Alex Cunningham
Frank Dellaert
Viorela Ila
Chris Beall
Eohan George
Manohar Paluri
Richard Roberts
Brian Law