Dzuchun
95c50b88a2
v0.4
...
v0.4.0
- feature-locked modules requiring `nalgebra/alloc` feature.
2025-01-18 03:41:12 +02:00
Dzuchun
f9d6f7e950
Fixed nalgebra/alloc feature mishandling
2025-01-18 03:36:25 +02:00
Dzuchun
7d73347051
Added katex.html to package
...
since apparently `docs.rs` DOES BUILD THE DOCS, what a surprise
2025-01-17 03:07:37 +02:00
Dzuchun
d1c15484e5
Made README.md work in a cargo package
v0.3.7
2025-01-17 02:59:47 +02:00
Dzuchun
f6c859e2e2
Excluded unnecessary files from package
2025-01-17 02:58:53 +02:00
Dzuchun
af4b7e1dd8
Version bump to 0.3.7
...
also:
- removed unnecessary dependencies
- now all dependencies use `crates.io`-provided versions
2025-01-17 02:58:53 +02:00
Dzuchun
9158a2f321
Added meta necessary to publish
2025-01-17 02:58:08 +02:00
Dzuchun
154e2807bc
Doc typos fix
2025-01-17 01:45:14 +02:00
Dzuchun
031e65c54c
License update
2025-01-17 00:10:01 +02:00
Dzuchun
36fdf78718
derive: Added meta for publication
2025-01-17 00:04:55 +02:00
Dzuchun
f65dbd9238
derive: dependencies bump
2025-01-17 00:04:08 +02:00
Dzuchun
50a1e5bf23
Updated miri action to a new syntax
v0.3.6
2025-01-04 14:05:05 +02:00
Dzuchun
86a2e79b81
[Cargo.lock]
2025-01-04 14:04:55 +02:00
Dzuchun
64e88b81b6
Proper crate version bumps
2025-01-04 14:04:55 +02:00
Dzuchun
8254510cb7
Made clippy happier
2025-01-04 14:04:55 +02:00
Dzuchun
4c15017270
Removed the tiny bit of unsafe present
...
So now there's no unsafe in this crate at all, and global
lint in workspace-level Cargo.toml strictly forbids it.
2025-01-04 14:04:55 +02:00
Dzuchun
d66537dbec
Upgrade compatible
2025-01-04 13:46:56 +02:00
Dzuchun
b355c90f20
Moved tests into separate files
v0.3.5
2024-11-27 02:16:03 +02:00
Dzuchun
6d1f09ffe0
[Cargo.lock]
2024-11-27 01:55:14 +02:00
Dzuchun
2bcf75d2ae
Massive Gaussian functions rework.
...
- There are asymmetric gaussian models now.
- Now Gaussian model can be fit for both sigma and s_p in any
combination.
- Renamed confusing `s` parameter to `sigma`.
- `Gaussian` and `GaussianW` are the same type now, whether
fit happens for `sigma` is determined by a generic const.
2024-11-27 01:51:48 +02:00
Dzuchun
d79c9993e4
Made &mut mandatory in fit/fit_stat
v0.3.4
2024-11-25 13:29:46 +02:00
Dzuchun
eb8d1798af
Fixed "model is still borrowed" bug
v0.3.3
2024-11-25 09:14:22 +02:00
Dzuchun
658b707e88
Convertion between identical FitStats
v0.3.2
2024-11-25 08:32:49 +02:00
Dzuchun
151f38010a
Simplified FitBound/FitBoundErr
...
v0.3.1
Now there's no need to specify `Y`-data container, if it's the same as
`X`-data container.
2024-11-25 05:57:51 +02:00
Dzuchun
e554a98e33
Renames of entity fields
2024-11-25 05:51:48 +02:00
Dzuchun
451a7644d7
Even more aggressive bound simplfication
2024-11-25 05:51:18 +02:00
Dzuchun
3a71ebfc4d
One more large type substitution
2024-11-24 17:58:31 +02:00
Dzuchun
b20fdcd0d3
Renamed Entity to Model in a bunch of places
...
v0.3
(leftovers from earlier development stages)
2024-11-24 17:15:57 +02:00
Dzuchun
9d4c06a9e9
Scalar is now an associated type of FitModel
...
(that's a massive rewrite!)
This change might seem random, but it, in fact, simplifies usage A LOT.
Now, `Scalar` type is **determined** by model, which is what should
happen in the first place.
2024-11-24 17:11:35 +02:00
Dzuchun
68641ac7e5
Hidden polynomial module from doc
2024-11-24 16:54:17 +02:00
Dzuchun
fe83a4e027
[Cargo.toml]
v0.2
2024-11-23 16:32:43 +02:00
Dzuchun
98760e9277
Fixed typos action
2024-11-23 16:26:06 +02:00
Dzuchun
dd5105d226
Restructured the crate
...
- Now using `generic_array_storage` v0.1.1 (Conv trait drastically
changed).
- `FitModel::ParamCount` type now has a less strict type bound, allowing
use of `typenum::Const` and even `nalgebra::Const`, if you feel like
it.
- Removed `Params` argument from `ConstOptimizationProblem` and
`DynOptimizationProblem`.
2024-11-23 16:26:06 +02:00
Dzuchun
b9e101e077
Copied simpliest macro impl to a separate file
...
For easier macro output testing
2024-11-23 16:26:01 +02:00
Dzuchun
c7c200491f
Removed mandatory dependency on libm over std
v0.1.1
2024-11-22 02:35:31 +02:00
Dzuchun
66f8a8ce5c
GenericArray is now a valid data source
2024-11-22 02:35:23 +02:00
Dzuchun
60c7acb009
Documented "reserved" name for the weights
2024-11-22 01:50:52 +02:00
Dzuchun
3d253ff023
FitModelErrors fix for Polynomial
v0.1
2024-11-22 00:35:50 +02:00
Dzuchun
a090595138
Tagged generic_array_storage
2024-11-21 23:56:45 +02:00
Dzuchun
88f733162c
Ignored some tests for Miri
...
- model jacobian test, because of stacked borrows issue inside ViewStorage (help wanted!?)
- stddev test, as it took too long
2024-11-21 23:39:38 +02:00
Dzuchun
1d2f0b8b9f
[minor] double colon in front of core
2024-11-21 23:35:34 +02:00
Dzuchun
815b850064
[Cargo.lock]
2024-11-21 21:18:36 +02:00
Dzuchun
103b310180
Added macro to generate test for model's jacobian
2024-11-21 21:15:36 +02:00
Dzuchun
0c682848f0
Fixed error computation & added test to enforce it
...
Note that error formula is an approximation for case of large datapoint count, so it's strict enforcement is not possible, unfortunately.
2024-11-21 21:15:36 +02:00
Dzuchun
c1e0b1308a
Ranged: changed parameters order
2024-11-21 21:15:36 +02:00
Dzuchun
5b350556e2
[minor] lint attribute in test
2024-11-21 21:15:36 +02:00
Dzuchun
b9964d5c80
Polynomial now uses IntoArrayLength
2024-11-21 21:15:36 +02:00
Dzuchun
8e54cd7325
Gaussian impl bug fixes
2024-11-21 21:15:36 +02:00
Dzuchun
0de471fd76
Composition model impl format
2024-11-21 21:15:36 +02:00
Dzuchun
b9df9653a6
DifferentiableFunction and ModelMap re-design
2024-11-21 21:15:36 +02:00