d02b5c4890
Cherry-picked: Added list_of.hpp dependence and fixed header order in doing so...
2014-09-23 11:42:53 +02:00
4a1d04a4d0
Applying patch from https://bitbucket.org/eigen/eigen/commits/d336163ce058/ to fix serialization on Mac with Boost 1.56.
...
https://forum.kde.org/viewtopic.php?f=74&t=122194
Matlab still crashes, however.
2014-09-22 16:51:09 -04:00
bc5f1c4a46
upgrade to Eigen 3.2.2
2014-09-22 16:20:53 -04:00
b47837462e
Unit test succeeds !!!
2014-09-21 18:54:01 +02:00
4bc82da85c
Compile with optional derivatives
2014-09-21 18:36:19 +02:00
8e3a0f4847
Prototype [jacobians] code
2014-09-21 18:22:28 +02:00
583c81ffea
Implemented uncalibrate, value test succeeds
2014-09-21 17:59:34 +02:00
6a5e4191a3
UnaryExpression
2014-09-21 17:43:47 +02:00
b89f65cccc
BinaryExpression
2014-09-21 17:37:09 +02:00
59b0e6a657
Progress on error
2014-09-21 17:13:25 +02:00
d9fafc1bf1
No more base class
2014-09-21 16:48:24 +02:00
7a64e2e546
Class hierarchy
2014-09-21 16:30:30 +02:00
20b9c31465
The range adaptor scheme did not work for std::map TERMS in creating a JacobianFacor. Hence, I removed it - which I think is more readable - and replaced it with an explicit creationg of dimensions. I also added a test with std::map, which works.
2014-09-21 13:58:45 +02:00
d34ba9c8bf
The range adaptor scheme did not work for std::map TERMS in creating a JacobianFacor. Hence, I removed it - which I think is more readable - and replaced it with an explicit creationg of dimensions. I also added a test with std::map, which works.
2014-09-21 13:40:38 +02:00
81645cb1f4
Added list_of.hpp dependence and fixed header order in doing so...
2014-09-21 13:38:12 +02:00
6532966f62
Call JacobianFactor constructor with map
2014-09-21 13:11:55 +02:00
c7b6a9af12
Now use old-style factor to create expected value and derivatives
2014-09-21 05:33:53 -04:00
a76c27d074
Now just linearize
2014-09-18 18:33:11 -04:00
b19132e004
First version of test, with vector of Matrices
2014-09-18 17:10:39 -05:00
3046d52d6e
Fix bugs in metis re building on Windows. Mostly version issues were new c math functions were redefined poorly in gklib and metis.
2014-09-18 10:45:01 -04:00
f9f6fadba4
Updating code documentation per issue #114
2014-09-15 11:52:40 -04:00
be68cc0047
add given marginals to update_noiseModel to improve speed
2014-09-15 03:03:25 -04:00
43a8de01c1
Merge remote-tracking branch 'origin/feature/quadratic_programming' into feature/quadratic_programming
...
Conflicts:
.cproject
.gitignore
gtsam/linear/GaussianFactorGraph.cpp
gtsam/linear/GaussianFactorGraph.h
gtsam/linear/HessianFactor.cpp
gtsam/linear/JacobianFactor.h
gtsam/linear/QPSolver.cpp
gtsam/linear/tests/testQPSolver.cpp
gtsam/nonlinear/NonlinearFactor.h
2014-09-14 01:45:27 -04:00
1a7653730c
fix headers for TransBtwRobotFactorEM
2014-09-13 18:12:28 -04:00
df1eede30c
Produce a graph of dual-scaled Hessians of each factor: -lambda*H, used for solving nonlinear equality constraints with SQP.
2014-09-13 01:37:17 -04:00
20fb8ab77d
Build a dual graph to compute dual values for equality constrained factors
2014-09-13 01:36:04 -04:00
4ca9d5757f
Support optional dual key for constrained Jacobian and nonlinear factors. Default boost::none for unconstrained factors.
2014-09-13 01:34:33 -04:00
5e8c36b3ca
compute gradient wrt a key
2014-09-13 01:31:49 -04:00
0dcb3e209d
add macro TEST_DISABLED for disabling tests without commenting them out
2014-09-13 01:29:25 -04:00
3ad83e6394
small improvements
2014-09-11 20:54:46 -04:00
6d697f2c92
heuristic in EliminationPreferCholesky to work around the Indeterminant exception while solving linear constrained systems.
...
Instead of turning Hessian factors into Jacobian factors -- so that they can be eliminated with constrained Jacobian factors using the special QR in Constrained's noise model -- we combine all Hessian factors, eliminate the variable first to have a conditional and a new factor 1, then combine the constrained Jacobians with this conditional (also a Jacobian) to eliminate again, producing the final conditional, and a new factor 2. The two new factors are then combined into a new Hessian factor to be returned.
2014-09-11 14:51:35 -04:00
661862f3e3
small comment
2014-09-11 12:08:42 -04:00
7322a74bbd
nice and clean unit tests for gradient initialization
2014-09-11 12:05:36 -04:00
db411214ff
renamed GPS factor
2014-09-10 15:30:15 -04:00
3d06a737bf
added possibility to specify number of iterations
2014-09-10 09:35:49 -04:00
6b890cec0e
added updateNoiseModels functionality
2014-09-10 00:24:32 +03:00
70956bb447
minor fixes for clang in C++11 mode on Mac
2014-09-08 15:52:07 -04:00
5bc4810dcb
Correcting bug fixes for Visual Studio. See bb issues #115,116,118 for more information
2014-09-08 15:19:54 -04:00
e3ec28ebeb
added debug info for gradient code
2014-09-06 18:27:07 -04:00
41bb99b48a
fixed gradient
2014-09-06 10:57:22 -04:00
aacd3484c7
bug fix: different pose covariance structure between g2o (t,R) and gtsam (R,t)
2014-09-04 10:07:28 -04:00
8444680e6e
fixed unit test
2014-09-03 20:20:06 -04:00
08a6bff803
fixed issues in dataset and added comments
2014-09-03 20:18:46 -04:00
544e56f634
added examples
2014-09-03 15:40:57 -04:00
316ce41a22
removed comments
2014-09-03 15:31:30 -04:00
b19e758f5a
no test failures
2014-09-03 15:30:41 -04:00
103d2a8ae9
working unit tests
2014-09-03 15:23:35 -04:00
35d5b56b65
still debugging
2014-09-02 20:56:36 -04:00
8384c37179
fixed 1 unit test
2014-09-02 20:17:31 -04:00
366101176c
added initial failing unit test
2014-09-02 18:50:23 -04:00
dellaert
cbeall3
Andrew Melim
Jing Dong
thduynguyen
Luca
Vadim Indelman