fly316/PX4-Autopilot
7
0
Fork 0
mirror of https://gitee.com/mirrors_PX4/PX4-Autopilot.git synced 2026-10-16 11:33:34 +08:00
Code Issues Packages Projects Releases Wiki Activity
397 Commits 852 Branches 161 Tags
1d0e7f1ca1187c90c70a9dac92ca6294a320b1d3
Commit Graph
397 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
Daniel Agar 1d0e7f1ca1 Matrix: use naive per element copy instead of memcpy call 2021-04-14 15:44:18 -04:00
Daniel AgarandMatthias Grob 1344ce06fd Euler: simplify DCM constructor 
Co-authored-by: Matthias Grob <maetugr@gmail.com>
 2021-04-14 15:40:24 -04:00
Daniel Agar 3679f7fd51 add double precision alias declarations and remove typedefs 2021-02-11 12:40:54 -05:00
Julian Kent 977cf52322 Ignore debug line in code coverage 2021-01-22 17:30:00 +01:00
Julian Kent 4837316f9b Fix small inconsistencies and compiler warnings 2021-01-22 17:30:00 +01:00
Julian Kent d540ca5de2 Coerce default epsilon values to Type 2021-01-13 10:17:09 +01:00
Julian Kent 054f8b12f4 Use a single inverse implementation for a single matrix size 2021-01-05 09:02:09 +01:00
Julian Kent 15e54ceda1 Rework rank-detection tolerance in pseudoinverse 2020-12-18 12:07:19 +01:00
Julian Kent d9a5e3dec2 Add test for new cutoff threshold 2020-12-18 12:07:19 +01:00
JacobCrabill ea43303c1c test/vector3: Add test for scalar +/- operators 2020-12-17 18:05:46 +01:00
JacobCrabill 158409abce Vector3f: Add operator+/- for scalars 2020-12-17 18:05:46 +01:00
Mitchell-Lee-93 a504b6e881 precision of tol for float has been changed 2020-11-04 10:00:55 +01:00
Julian Oes ce6b10b99a Fix clang-tidy warnings 
This comes from PX4 CI.
 2020-08-27 17:42:42 +02:00
Kamil Ritz 13f092a30f sparse quadratic form 2020-08-26 19:45:05 +02:00
Julian Oes 3a5bfb2bd1 matrix: inline to prevent multiple definitions (#147) 
Without inline we get multiple definitions during linking.
 2020-08-25 12:58:04 +02:00
Julian Oes e101edc0e7 Matrix: fix warning if M == N (#146) 
This fixes the warning appearing with GCC 10.2:

../../src/lib/matrix/matrix/Matrix.hpp:481:34: error: logical ‘and’ of equal expressions [-Werror=logical-op]
  481 |         for (size_t i = 0; i < M && i < N; i++) {
      |

I would prefered something with if constexpr but we don't have that yet
in because we're using C++14.
 2020-08-24 14:19:43 +02:00
Morten Fyhn Amundsen cd8ad1584c Make wrap() work with integer types (#145) 
* Make wrap() work with integers
 2020-08-24 10:49:12 +02:00
kamilritz 25c0455348 comply with Firmware style requirements 2020-08-10 10:13:40 +02:00
kritz f981cea2ae add possibility to iterate over SparseVector data at runtime (#143) 2020-08-09 09:44:49 +02:00
kamilritz e714a28c83 use size_t instead of int in sparseVector 2020-08-07 19:45:02 +02:00
kritzandJulian Kent 18699416b0 SparseVector (#140) 
* Add SparseVector

temp

* Add gtest

* Some reworking of the sparse concept

* Change type of M from int to size_t

* Add const modifier

* Add needed declaration for accessing elements of _indices

* Add norm_squared, norm, longerThan

* Add test for all sparse vector functions

* Add missing const to slice's norm_squared, norm and longerThan

* Construction from Vector<M> and carray[N]

* try to fix ci

Co-authored-by: Julian Kent <julian@auterion.com>
 2020-08-07 18:44:48 +02:00
kamilritz 6ed5dbc2db Construct Vector from Slice<1,M> alias row() 2020-08-06 16:15:05 +02:00
Matthias Grob a126be0882 attitude test: refactoring to avoid identity quaternion confusion 2020-07-20 18:44:08 +02:00
Matthias Grob 572bafb01f Quaternion: use template type for division by two 2020-07-20 18:44:08 +02:00
Matthias Grob 242b38ee9e Quaternion: remove deprecated axis_angle conversions 2020-07-20 18:44:08 +02:00
Matthias Grob 7a3009f7a3 Remove asserts for unsigned integer >= 0 
GCC 10 gives a warning
"comparison of unsigned expression in ‘>= 0’ is always true"
for these asserts since checking if an unsigned integer cannot
be negative and hence the statement gets droped.
 2020-07-20 15:10:49 +02:00
Matthias Grob 0fd99c59f1 Switch operator() to return a const reference 2020-07-17 10:43:34 +02:00
kamilritz 8a59b463f2 Quaternion: Use template type instead of single hardcoded type 2020-07-17 10:37:18 +02:00
Julian Kent 9a30828a0a Add explicit matrix + scalar test 2020-06-30 10:37:10 -04:00
Julian Kent f3cf615023 Do += -= and scalar *= /= in place 2020-06-30 10:37:10 -04:00
kamilritz f529358e9a Add missing const modifier 2020-06-28 14:24:49 +02:00
kamilritz 674bd99f3b Add operator* and operator/ for slice with type 2020-06-12 11:34:09 +02:00
Nicolas MARTIN 38b3acce96 fix conjugate_inversed comment 2020-06-11 16:17:28 +02:00
Matthias Grob 2bee0d078c Quaternion: refactor multiplication to matrix multiplication style 
Most often the multiplication in convention descriptions and papers is
described in matrix multiplication style like this:
q · p := Q(q)p

Q(q) :=
[q0 −q1 −q2 −q3]
[q1 q0 −q3 q2]
[q2 q3 q0 −q1]
[q3 −q2 q1 q0]

I'm just rearanging the terms such that it's easily comparable with
these definitions additional to it being clearly described by
documenting we use the hamilton convention.
 2020-05-06 13:17:02 +02:00
kamilritz d613055462 Add more assignment operators for slices 2020-04-07 09:12:17 +02:00
Matthias Grob 976ada411b Matrix: min max comments and test style 2020-03-23 09:46:07 +01:00
Matthias Grob fa7153ecfb Matrix: omit min max nan case with same result 
I observed this during review in
https://github.com/PX4/Matrix/pull/105/files#r348386226
 2020-03-23 09:46:07 +01:00
Matthias Grob a32892926c Matrix: capture stdout for print() test 
I want to switch the print() function back to use printf()
such that it's fully compatible with all (embedded) platforms.

To still test the print function I'm capturing stdout into a file
such that the print() function can stay unchanged and the result of
the printf()s can be evaluated.
 2020-03-12 13:24:08 +01:00
Matthias Grob c6db357c92 Matrix: correct print buffer size prediction 
Apparently printf of %8.8g can result in
8 significant digits + dot + scientific notation e.g. "e+12" = 14 bytes
 2020-03-12 10:08:04 +01:00
Matthias Grob 2cca35c8fe AxisAngle: call q.imag().norm() for conversion 
This is important depending on the platform because norm()
calls matrix::squrt() which dispatches correctly to the implementation
for the used type. Otherwise float squrt()s can get calculated as
double.
 2020-03-12 07:35:44 +01:00
kritz 4873dc1c1e Analytic inverse implementation (#122) 
* Add analytic 2x2 matrix inverse

* Add analytical 3x3 matrix inverse
 2020-03-04 09:14:04 +01:00
Julian Kent 649c837b6b Use faster but less accurate average 2020-02-25 15:08:17 +01:00
kamilritz a37b91c96a Type cast remaining integer 2020-01-13 12:21:18 +01:00
Lorenz Meier 3b581fb599 Update .travis.yml 
Bring Ubuntu and Mac OS toolchains to latest.
 2020-01-01 15:53:50 +01:00
Lorenz Meier efa5580a67 Update CMakeLists.txt 
Remove non-compatible compile option.
 2020-01-01 15:53:50 +01:00
Lorenz Meier 20a9e91d44 Fix CMake configuration for coverage tests on Mac OS 2019-12-31 15:54:51 +01:00
kritz d18be0d0fa Fix AngleAxis constructors (#117) 2019-12-28 17:18:15 -05:00
kritz e81483a808 Catch quaternion canonical corner cases (#116) 2019-12-18 14:12:47 +01:00
kritz 4f3565da94 Add asserts (#115) 
* Add asserts

* Type cast literals

* asserts for indexing vectors

* include assert

* Fix accessing elements outside of slice
 2019-12-09 10:21:27 +01:00
kritz 2f6398168d Add several functions that are useful for coavriance matrices: (#114) 
* uncorrelateCovariance
	* uncorrelateCovarianceSetVariance
	* makeBlockSymmetric
	* makeRowColSymmetric
	* isBlockSymmetric
	* isRowColSymmetric
 2019-12-06 12:03:26 +01:00