Commit Graph

1437 Commits

Author SHA1 Message Date
metric-space 6ac6e7f75e First compiling commit for take-2 of polar-decomposition:
Code inspired by this thread: https://github.com/dimforge/nalgebra/pull/656
Main person behind this is LucasCampos
2021-12-22 00:12:27 -05:00
haibane_tenshi 24bb7bef76 fix(Matrix): relax T: Zero bound on Matrix::select_rows() and Matrix::select_columns() 2021-12-16 18:50:15 +03:00
Sébastien Crozet e0a1b1bc34 Fix the special-case for 3x3 Real SVD 2021-12-09 11:52:37 +01:00
Sébastien Crozet 9389cf2adc
Merge pull request #1034 from dimforge/specific_svd
Add dedicated implementations of SVD for 2x2 and 3x3 real matrices.
2021-12-01 13:55:10 +01:00
Sébastien Crozet 9297cc5754 Implement DeviceCopy for UnitComplex, UnitQuaternion, and Unit<Matrix> instead of using a blanket impl 2021-11-26 18:13:37 +01:00
Sébastien Crozet 49e9ceea30 Add dedicated implementations of SVD for 2x2 and 3x3 real matrices. 2021-11-26 17:45:42 +01:00
Sébastien Crozet 9d897a6b8a Add support for rust-cuda 2021-11-23 12:03:11 +01:00
Sébastien Crozet ad3eefe182 Merge remote-tracking branch 'origin/dev' into dev 2021-11-23 11:02:51 +01:00
Sébastien Crozet f715883f9f Re-add the conversion from a slice to a static array 2021-11-21 21:47:35 +01:00
Sébastien Crozet 10150ec783
Merge pull request #1016 from tpdickso/fix-dual-quaternion-sclerp
Don't panic ScLERPing `UnitDualQuaternion` with equal rotation
2021-11-21 17:57:34 +01:00
Terence 0ecbed512b cargo fmt 2021-11-20 09:12:45 -05:00
Christopher Gundler 24d29c4de3 Allow sorting SVD according to singular values 2021-11-08 10:27:53 +01:00
Sébastien Crozet df51de0664 Fix unsound usize::get_unchecked_mut 2021-10-31 14:27:32 +01:00
Sébastien Crozet 20d8e953c5
Merge pull request #1004 from MaxVerevkin/fix-750
fix linear indexing of matrix slices
2021-10-31 14:24:10 +01:00
Terence 3df81c7cc9 fix docs 2021-10-28 00:05:50 -04:00
Terence dfb7b6af22 Don't panic ScLERPing `UnitDualQuaternion` with equal rotation
Solves #1013.

Previously, when screw-linearly interpolating two unit dual quaternions
that had an identical orientation, `try_sclerp` would return `None`, as
the operation would introduce a division-by-zero.

This PR splits out the cases where two unit dual quaternions have an
identical orientation from the cases where they have opposite
orientations. In the case where they have identical orientations, the
operation is well-defined, but the exponential parameterization could
not handle it without introducing NaNs. Therefore, the function detects
this case and simply defaults to linearly interpolating the
translational components and using one of the two inputs' rotation
components.

The case where the inputs have opposite rotations is now detected
separately using the dot product of the real (rotation) parts, which was
already being computed anyway.

Also introduces proptests for these specific scenarios, to avoid any
regression.
2021-10-28 00:02:20 -04:00
Sébastien Crozet e05bfe48b3
Merge pull request #1012 from Yuri6037/scale
Scale
2021-10-25 10:20:07 +02:00
Sébastien Crozet 4be7db36fd
Remove useless `into_owned`. 2021-10-25 09:57:58 +02:00
Yuri Edward 209f77ccaf Fixed from_superset_unchecked 2021-10-24 22:26:01 +02:00
Yuri Edward 25dba52197 Removed additional into_owned call 2021-10-24 21:58:05 +02:00
Yuri Edward 54b8a5b900 Rollbacked buggy update 2021-10-24 21:55:11 +02:00
Yuri Edward 50f65123b5 Merge branch 'scale' of github.com:Yuri6037/nalgebra into scale 2021-10-24 21:52:45 +02:00
Yuri6037 b8800ccd15
Update src/geometry/scale_conversion.rs
Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 21:52:35 +02:00
Yuri Edward 9d2639b742 Removed clippy suspicious_arithmetic_impl allow 2021-10-24 21:51:36 +02:00
Yuri Edward 866cabaa28 Merge remote-tracking branch 'origin/scale' into scale 2021-10-24 20:55:35 +02:00
Yuri Edward 6d3978b212 Added Scale *= scalar + reordered ops 2021-10-24 20:55:10 +02:00
Yuri Edward 0aba533556 Added Scale * Vector 2021-10-24 20:45:40 +02:00
Yuri6037 8efa9f76b7
Update src/geometry/scale_coordinates.rs
Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 20:41:27 +02:00
Yuri6037 510ffcbb18
Update src/geometry/scale_coordinates.rs
Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 20:41:21 +02:00
Yuri Edward c11e1ffb4d Fixed missing curly bracket 2021-10-24 20:40:02 +02:00
Yuri6037 ded853757c
Update src/geometry/scale.rs
Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 20:31:35 +02:00
Yuri6037 f7eec7317c
Update src/geometry/scale_conversion.rs
Removed return.

Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 20:27:54 +02:00
Yuri6037 5b44502632
Update src/geometry/scale.rs
Simplified code.

Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 20:27:38 +02:00
Yuri6037 26e36c15ba
Update src/geometry/scale.rs
Removed return

Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 20:25:48 +02:00
Yuri6037 6123c86f9e
Update src/geometry/scale.rs
Removed return

Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 20:24:16 +02:00
Yuri6037 91b8424a86
Update src/geometry/scale.rs
Removed return

Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 20:21:07 +02:00
Yuri6037 de41c42af7
Update src/geometry/scale.rs
Simplified code

Co-authored-by: Sébastien Crozet <sebastien@crozet.re>
2021-10-24 20:20:13 +02:00
Yuri Edward b91ead7971 Fixed Subset for Matrix 2021-10-21 15:56:07 +02:00
Yuri Edward 4466fb2514 Added point * vector 2021-10-21 15:47:00 +02:00
Yuri Edward 99704a0ea1 Updated all examples related to inversion API 2021-10-21 15:22:18 +02:00
Yuri Edward 82b8ac84a0 Fixed code formatting 2021-10-21 15:10:26 +02:00
Yuri Edward a64417870b Changed inverse API to pseudo_inverse, try_inverse, try_inverse_mut, try_inverse_transform_point, inverse_unchecked. 2021-10-21 15:06:25 +02:00
Yuri Edward 1341801c56 Added scale * scalar + Cleanup scale_ops 2021-10-19 16:33:12 +02:00
Yuri Edward 865ac001d3 Re-armed scale_simba 2021-10-19 16:20:25 +02:00
Yuri Edward 94ef916f49 Fixed cargo fmt 2021-10-19 16:18:25 +02:00
Yuri Edward 0e2865085e Updated doc tests 2021-10-19 16:09:06 +02:00
Yuri Edward d82b39ca9a Added to_homogeneous 2021-10-19 16:01:02 +02:00
Yuri Edward f1f947b924 Added to_homogeneous 2021-10-19 15:55:06 +02:00
Yuri Edward c249cc76b0 Added inverse, inverse_mut inverse_transform_point and transform_point 2021-10-19 15:02:50 +02:00
Yuri Edward cf811abb92 Added identity() 2021-10-19 14:47:46 +02:00
Yuri Edward bfb4aceede Updated operator overloads + fixed scale_construction 2021-10-19 14:31:20 +02:00
Yuri Edward 9c506b18e1 Renamed all instances of Translation to Scale 2021-10-19 14:21:26 +02:00
Yuri Edward 1252fb00a0 Added scale.rs 2021-10-19 14:05:51 +02:00
Yuri Edward 8c6ad490bc Initial commit copy all translation_x files into scale_x files 2021-10-19 14:02:07 +02:00
Sébastien Crozet ec700ab7cf Fix conflicting Default impl for unit-quaternions. 2021-10-14 11:48:12 +02:00
Sébastien Crozet 7275b5146e Implement Default for most geometry types 2021-10-14 11:39:22 +02:00
Max Verevkin bf48e093a9 fix linear indexing of matrix slices 2021-10-06 18:02:02 +03:00
Sébastien Crozet 97a79a994b Add methods to compute the products of a single matrix components/rows/columns 2021-10-06 10:48:46 +02:00
Sébastien Crozet 2e9c8aef20
Merge pull request #1002 from MaxVerevkin/unsafe-dim
make Dim trait unsafe
2021-10-04 09:32:13 +02:00
Max Verevkin 65d3b42ff9 make Dim trait unsafe 2021-10-03 20:26:34 +03:00
Max Verevkin ba7ededbf9 remove unnecessary `T: Scalar` 2021-09-26 21:07:34 +03:00
Max Verevkin 5cbff59f80 simplify Matrix::is_identity while also improving performance 2021-09-26 19:34:19 +03:00
Sébastien Crozet 7f236d88aa
Merge pull request #991 from MaxVerevkin/fix-ub
Fix UB in `Matrix::perp()`
2021-09-26 11:05:44 +02:00
Sébastien Crozet 5d9dbb62ef
Merge pull request #993 from MaxVerevkin/deny-unused-mut
deny unused_mut and unused_variables
2021-09-26 10:59:27 +02:00
Max Verevkin f4a8b4cfa1 deny unused_mut and unused_variables 2021-09-25 22:05:31 +03:00
Max Verevkin b91eecebcd Matrix::perp() fix UB 2021-09-25 21:52:09 +03:00
Sébastien Crozet 85202f85f7
Merge pull request #986 from Ralith/concise-debug
Concise Debug impls
2021-09-25 17:23:14 +02:00
Sébastien Crozet dd8b6800f5
Merge pull request #992 from MaxVerevkin/exp-rs
exp.rs: factorial(): use precomputed factorial array
2021-09-25 12:32:11 +02:00
Sébastien Crozet d12ad891f9
Merge pull request #981 from geieredgar/dev
Implement conversion traits for glam 0.18
2021-09-25 12:12:25 +02:00
Max Verevkin ad8250c361 exp.rs: factorial(): use precomputed factorial array 2021-09-21 23:14:42 +03:00
Sébastien Crozet 654eca7f80
Merge pull request #979 from Fuuzetsu/cholesky-lax
Allow setting Cholesky field directly; more lax decomposition method
2021-09-17 05:16:39 -07:00
Benjamin Saunders 0541f13b26 Concise Debug impls
Replace the verbose derived (or nearly equivalent) Debug impls for
several newtypes with explicit impls that forward to the inner type,
making readable diagnostics logging much easier.
2021-09-12 10:56:25 -07:00
Mateusz Kowalczyk fc3526b58a
Allow fallback value for Cholesky decomposition
This is useful in case where the values are very close to zero or
possibly even slightly negative. This can quite easily happen due to
numerical errors.

A common strategy is to replace these values with a small epsilon value
that keeps the matrix SPD. Some libraries even do this by default (such
as https://github.com/STOR-i/GaussianProcesses.jl/issues/1).

We point the user to `LU` decomposition and also make it clearer that
the method is basically a hack. The public method no longer takes an
`Option` which didn't really make sense. A private method is used to not
repeat implementation in `new`.
2021-09-03 09:01:04 +09:00
Mateusz Kowalczyk a4c2ca941d
Allow constructing Cholesky struct directly
This is useful if there the data for the matrix comes from elsewhere,
such as a custom algorithm. One such use-case can be seen
https://github.com/nestordemeure/friedrich/issues/43 where we do special
handling for the try_sqrt or is_zero cases that would normally fail the
decomposition in `Cholesky::new()`.
2021-09-03 09:01:04 +09:00
Edgar Geier bc21b6f1e6
Add support for conversion with glam 0.18 2021-08-30 18:31:21 +02:00
Sébastien Crozet 319597e137
Merge pull request #974 from cchillen/RotationDocumentation
Improve clarity of Rotation doc comments
2021-08-30 10:32:47 +02:00
Sébastien Crozet 90e531337a
Merge pull request #969 from remilauzier/dev
Fix some clippy warnings
2021-08-30 10:11:51 +02:00
Rémi Lauzier ea9a9e8b7f
Fix some clippy warnings 2021-08-28 12:05:21 -04:00
cchillen 5b84c46840 Update imports in doc example. 2021-08-28 09:41:09 -04:00
cchillen 25341e40f0 Improve clarity of Rotation doc comments
The doc comments for `Rotation` incorreclty refer to quaternion instead of a rotation matrix. No code change, purely documentation.
2021-08-28 09:09:53 -04:00
raimundo saona 2233a38048 Implement conversion methods for RowDVector 2021-08-26 11:07:00 +02:00
Sébastien Crozet bd9e0fb72c Add support for conversion with glam 0.16 and 0.17 2021-08-08 17:41:10 +02:00
Sébastien Crozet a9ed6cb621 Fix wrong sign in UnitComplex::axis_angle 2021-08-08 13:23:09 +02:00
Sébastien Crozet 1bc919e0db
Merge pull request #962 from dimforge/non-copy-types
Tha clone wars: allow non-copy scalar types everywhere
2021-08-08 13:21:39 +02:00
Marius Knaust 649e0f5a59
Fix wrong reference in docs. 2021-08-07 12:06:15 +02:00
Sébastien Crozet dd6c40016e Remove the Copy requirement from SimdRealField. 2021-08-04 17:34:25 +02:00
Sébastien Crozet 65b299557c More inlining. 2021-08-04 11:19:57 +02:00
Sébastien Crozet 6d57396a42 Remove the Scalar::is method, which is unsound. 2021-08-03 17:53:48 +02:00
Sébastien Crozet eedb860565 Fix missing docs. 2021-08-03 17:27:15 +02:00
Sébastien Crozet d609a2f174 Address unsoundness in the resizing API. 2021-08-03 17:02:42 +02:00
Sébastien Crozet 27ae30b46a First step to fix unsoundness on the resize API. 2021-08-03 10:59:03 +02:00
Sébastien Crozet f67a81e50a Fix build with --all-features 2021-08-03 10:07:05 +02:00
Sébastien Crozet 7a1a4bcc02 Fix test and no-std builds. 2021-08-03 09:54:11 +02:00
Sébastien Crozet 8c6ebf2757 Implement the single-allocator-trait approach. 2021-08-02 18:41:46 +02:00
Sébastien Crozet d34fed45bc Merge branch 'dev' into ub3 2021-08-02 18:40:33 +02:00
Sébastien Crozet 441bfcb3ff
Merge pull request #955 from qu1x/reflection
Add bias getter and alias for reflection.
2021-08-01 10:29:11 +02:00
Violeta Hernández 2243a11e89 Added some derives 2021-07-29 12:33:45 -05:00
Rémi Lauzier c35f792b4f
Fix some warnings 2021-07-27 19:24:55 -04:00
Sébastien Crozet 2a80e96766
Merge pull request #948 from CAD97/ops-transform-complex
Add Transform × UnitComplex & friends
2021-07-27 15:22:53 +02:00
Sébastien Crozet 99a79d51f5
Merge pull request #925 from CAD97/bytemuck
Impl bytemuck traits for geometry types
2021-07-27 15:19:14 +02:00
Sébastien Crozet 15c4a25c96
Merge pull request #950 from CAD97/moar-const
Make `from_matrix_unchecked`es const
2021-07-27 15:18:42 +02:00
Rouven Spreckels 9824fbc67b Add reflection alias. 2021-07-27 15:18:07 +02:00
Rouven Spreckels 06e20b4b95 Add getter for reflection bias. Fix typos. 2021-07-27 15:17:30 +02:00
Sébastien Crozet 0d3af06e01
Merge pull request #952 from remilauzier/dev
Fix rust_2018_idioms warnings
2021-07-27 10:29:40 +02:00
Andreas Longva de540aa7ff Do not require T: (De)Serialize for OPoint impl
The bounds recently got a little too strict by accident.
2021-07-26 17:57:24 +02:00
Rémi Lauzier ed6b34a0d6
Fix rust_2018_idioms warnings 2021-07-25 13:18:21 -04:00
CAD97 ceb30a68b8 Fix a few bitrotted user guide links 2021-07-23 21:22:59 -05:00
CAD97 04dd3ff241 Make some from_matrix_unchecked const 2021-07-22 18:07:12 -05:00
Christopher Durham 07c3fbc191
Merge branch 'dev' into bytemuck 2021-07-22 17:47:57 -05:00
Violeta Hernández 22b657f566 First draft of `Owned` overhaul 2021-07-19 18:00:40 -05:00
Violeta Hernández a753d84aae cargo fmt 2021-07-17 21:50:06 -05:00
Violeta Hernández b74be8499f Miscellaneous improvements 2021-07-17 21:43:50 -05:00
Violeta Hernández 10b5dc9bb6 Many miscellaneous improvements throughout 2021-07-17 20:19:20 -05:00
Violeta Hernández 7e1b2f81b3 Fixed some more blatant issues 2021-07-17 13:53:01 -05:00
Violeta Hernández fa1ed9683b Got rid of the `unimplemented_or_uninitialized_generic` macro 2021-07-17 13:36:35 -05:00
Violeta Hernández 0687318c7a Tests work! 2021-07-17 13:30:57 -05:00
Violeta Hernández c01d591478 We have reached compilation! 2021-07-17 13:01:03 -05:00
Violeta Hernández 9a528e23b9 Almost! 2021-07-17 04:36:14 -05:00
Violeta Hernández a6b8dd6d78 Checkpoint #10 2021-07-17 02:52:57 -05:00
Violeta Hernández 87fe2b30df Checkpoint #9 2021-07-16 23:17:56 -05:00
Violeta Hernández c3f869e017 Checkpoint #8 2021-07-16 01:53:28 -05:00
Violeta Hernández 8270dd8e89 `ops.rs` works too now! 2021-07-16 00:39:15 -05:00
Violeta Hernández 54e9750191 Tied some blas loose strings 2021-07-16 00:27:16 -05:00
Violeta Hernández df9b6f5f64 blas.rs works now! 2021-07-15 23:56:58 -05:00
CAD97 b1775ee747 Add Transform × UnitComplex & friends 2021-07-14 23:52:38 -05:00
Violeta Hernández bbd045d216 `blas.rs` should be sound now 2021-07-14 23:30:31 -05:00
Violeta Hernández 775917142b More trait restructuring! 2021-07-14 17:21:22 -05:00
Violeta Hernández 8d10e69e33 Finally figured out some trait nitty-gritty 2021-07-14 13:24:27 -05:00
Violeta Hernández 1a78b00476 Checkpoint #1 2021-07-14 04:25:16 -05:00
Sébastien Crozet cce7d9e9b6
Merge pull request #937 from jsmith628/fixed-size-borrow
Implemented Borrow and BorrowMut on fixed size matrices
2021-07-11 17:48:09 +02:00
Sébastien Crozet c320988088
Merge pull request #944 from vadixidav/dev
import macros from alloc
2021-07-11 17:47:51 +02:00
Sébastien Crozet b007e192d4
Merge pull request #938 from Nateckert/omatrix_to_string
Make display generical for Omatrix
2021-07-11 17:45:02 +02:00
Geordon Worley 02e9ac4cad import macros from alloc to solve issues with missing format! macro, among other issues 2021-07-10 18:33:11 -05:00
Sébastien Crozet 7bcb54641d
Merge pull request #943 from dimforge/point_dim_name
Add a OPoint type that takes type-level integers instead of const-generics
2021-07-10 11:40:31 +02:00
Sébastien Crozet ac61e112dd
Merge pull request #932 from CAD97/moar-const
Remove scalar bound from geometry type defs
2021-07-10 11:28:23 +02:00
Crozet Sébastien 2ce6811e2c Fix compilation when enabling the bytemuck feature. 2021-07-10 11:24:23 +02:00
CAD97 85d07b22a3 FIx serde impl bounds 2021-07-09 15:30:06 -05:00
CAD97 22ba88353a Fix bounds for bytemuck impls 2021-07-09 15:23:08 -05:00
Violeta Hernández 1be8964c12 Fixed indent 2021-07-08 11:15:08 -05:00
Violeta Hernández 04e937792b Removed unnecessary `unsafe` markers 2021-07-08 11:12:43 -05:00
Joshua Smith a7fe7a5d9a fixed check-fmt error 2021-07-08 11:05:32 -05:00
nathan.eckert 92a51662bf Remove Allocator<usize, > from display 2021-07-08 14:18:22 +02:00
Sébastien Crozet 0a5bd520a4
Merge pull request #922 from Cryptjar/impl-into-slice
Allow to extract the inner slices of SliceStorage
2021-07-08 09:50:41 +02:00
Sébastien Crozet a66b54a015
Merge pull request #929 from remilauzier/dev
Fix some clippy warnings and typo
2021-07-08 09:44:43 +02:00
Joshua Smith 2279da446e impl Borrow and BorrowMut for fixed size matrices 2021-07-07 13:26:19 -05:00
Violeta Hernández 6e96c28ad9 Trivial doc comment improvements 2021-07-06 21:05:25 -05:00
Rémi Lauzier bbd60fbc8c
Fix some clippy warnings and typo 2021-07-06 11:39:29 -04:00
Sébastien Crozet 1fc29dcf63
Merge pull request #930 from CAD97/patch-1
impl Hash for Transform
2021-07-06 01:09:00 -07:00
CAD97 314b4dd103 Remove scalar bound from geometry type defs
This was inconsistently applied, with some types having <T>, some having
<T: Scalar>, and some having <T: RealField>.
This unifies all types to match the convention of Matrix:
Just declare <T> at type def time, and apply bounds on impls only.

A significant advantage of this approach is const fn construction. Const
fn generics currently still can't have trait bounds, so any generic
const fn needs to only move opaque types around. Construction methods
such as new_unchecked or from_parts can be made const by removing their
generic bounds after this PR.

Actual constification is left to a follow-up PR.

Note that na::Transform is _not_ loosened here, as it has more complicated
definition requirements.
2021-07-05 19:04:10 -05:00
Christopher Durham 1eeaf3526c
Rotation::matrix_mut_unchecked may not cause UB
...because it isn't unsafe, so clarify the documentation to match.
2021-07-05 18:53:11 -05:00