Tensorwaves

_See all documentation for this version [here](https://tensorwaves.rtfd.io/en/0.3.0)._

💡 New features

<details>
<summary>Implemented FitResult.count_number_of_parameters (291 and 297)</summary>

Closes 290

Added a method `FitResult.count_number_of_parameters()`

</details>

<details>
<summary>Constructor arguments of FitResult are now validated (291)</summary>

</details>

<details>
<summary>Mapping of parameters values/errors in FitResult are now 'pretty printed' (291)</summary>

python
FitResult(
minimum_valid=True,
execution_time=2.9841978549957275,
function_calls=166,
estimator_value=-7579.247627538372,
parameter_values={

_See all documentation for this version [here](//tensorwaves.rtfd.io/en/0.2.9)._

🔨 Internal maintenance

- TensorWaves v0.2.9 can be installed with TensorFlow v2.6 and higher versions of [`phasespace`](https://github.com/zfit/phasespace) (#447)

_See all documentation for this version [here](https://tensorwaves.rtfd.io/en/0.2.8)._

💡 New features

<details>
<summary>A time estimate is now provided when generating data (286)</summary>

Fixes the `tqdm` progress bar so that it gives a time estimate when calling `generate_phsp` and `generate_data`.

</details>


🐛 Bug fixes

<details>
<summary>Analytic continuation can now be activated safely when using JAX as backend (284)</summary>

Implementation of [TR-001](https://compwa-org.readthedocs.io/report/001.html). Requires https://github.com/ComPWA/ampform/pull/75 and AmpForm v0.9.2 (upcoming).

</details>


🖱️ Developer Experience

<details>
<summary>Milestone are automatically closed when a matching release is published (282)</summary>

</details>

<details>
<summary>A new notebook show how to activate analytic continuation (285)</summary>

</details>

_See all documentation for this version [here](https://tensorwaves.rtfd.io/en/0.2.7)._

💡 New features

Larger SymPy expressions that take a long time to lambdify can now be lambdified faster (order of 10-100x) with `split_expression()`, `optimized_lambdify()`, and the `max_complexity` argument.
https://tensorwaves.rtfd.io/en/0.2.7/usage/faster-lambdify.html

_See all documentation for this version [here](https://tensorwaves.rtfd.io/en/0.2.6)._

💡 New features

<details>
<summary>Constructing an Estimator from a Function (272)</summary>

If the amplitude model is complicated, it can take over an hour to lambdify it to a `Model`. So it's annoying to have to do that twice if you use the same backend for data generation and fitting. This PR allows reusing a `LambdifiedFunction` for fitting.

</details>



<details>
<summary>Data can now be generated with JAX as well (271)</summary>

Simple fix: just casts [`ScalarSequence`](https://ampform.readthedocs.io/en/0.9.1/api/ampform.data.html#ampform.data.ScalarSequence) back into a `numpy` array so that a lambdified `Function` can handle it.

</details>

<details>
<summary>Python 3.9 and TensorFlow 2.5 are now supported (275)</summary>

Many thanks to mayou36 for releasing [`phasespace` v1.3](https://github.com/zfit/phasespace/releases/tag/1.3.0)!

</details>

_Added support for analytic continuation. See all documentation for this version [here](https://tensorwaves.rtfd.io/en/0.2.5)._

💡 New features

<details>
<summary>Upgrade to AmpForm v0.9.x (270)</summary>

Adds support for analytic continuation in particular.

</details>

🐛 Bug fixes

<details>
<summary>Lambdify to numpy, not numpy.__dict__ (f0cbcaa)</summary>

Lambdifying to with `modules=numpy.__dict__` does not work with conditionals (`sympy.Piecewise`). If using numpy.**dict**, Piecewise is lambdified to if-else (which doesn't worky with numpy arrays), while it should convert to `numpy.select`.

</details>

🖱️ Developer Experience

<details>
<summary>Run notebooks with pytest (982aa21)</summary>

</details>

<details>
<summary>Simplified usage examples to speed up CI (0737f49)</summary>

</details>