Foxes

- Command line applications:
- New command line application: `foxes_yaml`, runs *foxes* by interpreting a `yaml` input parameter file (no script neccessary)
- New command line application: `foxes_windio`, runs *foxes* by interpreting a windio `yaml` input parameter file (no script neccessary)
- Config:
- Introducing a dictionary-like configuration object `foxes.config`, setting default data types, the directory choices, etc.
- Inputs:
- New input package `foxes.input.yaml`, realizing the run via yaml parameter files. This is not following `windio`, but a *foxes* specific structure.
- Reworked `windio` input package, now at `foxes.input.yaml.windio` and based on `foxes.input.yaml`
- Outputs:
- Introducing the idea of a default output directory, accessible via `foxes.config.out_dir`. The file writing output classes now refer to this, or a output specific output directory that is given as an argument to the constructor of the output class.
- Examples:
- New example `yaml_input`, demonstrating the command line application `foxes_yaml` and the new yaml parameter file
- Bug fixes:
- Fix for `TurbOParkIX` wake model: Now `k` is truly `k`, such that only `ambka` or `ka` choices from `WakeK` correspond to multiplication with `AMB_TI` or `TI`, respectively.

**Full Changelog**: [https://github.com/FraunhoferIWES/foxes/commits/v1.2](https://github.com/FraunhoferIWES/foxes/commits/v1.2)

- Dependencies:
- Removing optional dependencies `io`. The installation advice will be printed when trying to use the `foxes.input.windio` sub package. The reason is that for now this depends on a personal [fork by kilojoules](https://github.com/kilojoules/windIO), and the inclusion of such dependencies is not supported by PyPi.

**Full Changelog**: [https://github.com/FraunhoferIWES/foxes/commits/v1.1.1](https://github.com/FraunhoferIWES/foxes/commits/v1.1.1)

- Python versions:
- Adding support for Python3.13
- Engines:
- Default number of processes is now os.cpu_count(), or os.process_cpu_count() for python3.13
- New engine: `RayEngine` (short name `ray`), runs parallel computations based on the [Ray package](https://docs.ray.io/en/latest/)
- Inputs:
- New states `SliceDataNC`, much like `FieldDataNC` but without height dependency
- Models:
- New turbine type `TBLFile`: Reads power, ct, D, H, P_rated from a *.tbl file
- Turbine induction models `SelfSimilar`, `SelfSimilar2020`, `Rathmann`, `VortexSheet` now optionally accept any wind speed superposition model, i.e., they are no longer based on hard-coded linear superposition
- Bug fixes:
- Bug fixed in `WSProduct`, causing zero wind speed at regions not touched by wakes
- Bugs fixes in `FarmLayoutOutput`, concerning the writing of the layout csv file

**Full Changelog**: [https://github.com/FraunhoferIWES/foxes/commits/v1.1](https://github.com/FraunhoferIWES/foxes/commits/v1.1)

This major version introduces the concept of `Engines` which handle the chunking and parallelization of all *foxes* calculations. The default choice now prefers the [concurrent.futures](https://docs.python.org/3/library/concurrent.futures.html) package and provides a significant speedup compared to previous versions. See the documentation for more details and all engine choices. The `Engines` replace the `Runners` of previous versions.

- Engines
- New engine `ThreadsEngine` (short `threads`): Sends chunks to threads, based on `concurrent.futures`
- New engine `ProcessEngine` (short `process`): Sends chunks to processes, based on `concurrent.futures`
- New engine `MultiprocessEngine` (short `multiprocess`): Sends chunks to a multiprocessing pool
- New engine `XArrayEngine` (short `xarray`): Runs parallelization via [xarray.apply_ufunc](https://docs.xarray.dev/en/stable/generated/xarray.apply_ufunc.html)
- New engine `DaskEngine` (short `dask`): Submits chunk calculation functions to `dask`
- New engine `LocalClusterEngine` (short `local_cluster`): Creates a virtual cluster on the local machine
- New engine `MPIEngine` (short `mpi`): Parallel computation via `mpi4py`, also multi-node
- New engine `SlurmClusterEngine` (short `slurm_cluster`): Submits jobs to a SLURM system
- New engine `NumpyEngine` (short `numpy`): Runs a loop over chunks
- New engine `SingleChunkEngine` (short `single`): Runs single-chunk calculations
- New engine `DefaultEngine` (short `default`): Switches between `single` and `process`, depending on the case size
- Inputs:
- New states `OnePointFlowStates`, `OnePointFlowTimeseries`, `OnePointFlowMultiHeightTimeseries`, `OnePointFlowMultiHeightNCTimeseries`: Generating horizontally inhomogeneous inflow from horizontally homogeneous input data
- New farm layout option: `add_ring`, adding a ring of turbines
- Models:
- Wake frame `Timelines` now also accept spatially uniform multi-height states
- New wake frame `DynamicWakes`: Dynamic wakes for any kind of timeseries states, compatible with chunking
- New turbine type `FromLookupTable`, computes power and thrust coefficient from a lookup table
- Outputs:
- New sub package `seq_plugins`, in case more of these will be added in the future
- New sequential plugin `SeqWakeDebugPlugin`, adding wake centres and velocity vectors to flow animations, for debugging
- Examples:
- New example: `dyn_wakes`, similar to `timelines` but with dynamic wakes and `OnePointFlowTimeseries` inflow

**Full Changelog**: [https://github.com/FraunhoferIWES/foxes/commits/v1.0](https://github.com/FraunhoferIWES/foxes/commits/v1.0)

- Outputs:
- Improved `SliceData` output: Now either specify `resolution` or `n_img_points`, e.g. `n_img_points=(100, 100)` for an image with 100 x 100 points

**Full Changelog**: [https://github.com/FraunhoferIWES/foxes/commits/v0.8.3](https://github.com/FraunhoferIWES/foxes/commits/v0.8.3)

- Removing `plotly_helpers.py` from `utils`
- Updating dependencies

**Full Changelog**: [https://github.com/FraunhoferIWES/foxes/commits/v0.8.2](https://github.com/FraunhoferIWES/foxes/commits/v0.8.2)