Simphony

Some developer updates are mainly included in this release, along with a switch
back to the MIT License (from the GPLv3+ License).

Added
- Pre-commit hooks added to assist with code quality.
- Bump2version (pip) now manages version numbering throughout the codebase.
- [42](https://github.com/BYUCamachoLab/simphony/pull/42) Added the ability to
simultaneously sweep over multiple inputs.

Changed
- Switched back to the MIT License from the GPLv3+.

Removed
- Some of the source files taken from the SiEPIC-Tools and SiEPIC EBeam PDK
packages, which were causing issues with the black codeformatter and were
unnecessary (as we only use the data files for the compact models anyway).
- The (unimplemented) classes `SinglePortSweepSimulation` and
`MultiInputSweepSimulation` were removed in favor of adding single/multi
input abilities to the base class `SweepSimulation`.

---

[0.3.0](https://github.com/BYUCamachoLab/simphony/tree/v0.3.0) - <small>2020-05-18</small>

This version is a complete codebase rewrite implementing a much more
human-friendly way of defining circuits and running simulations.

Circuits are now defined and stored in a clear, easy-to-understand
object-oriented way. Integration with other packages should be easy, and the
creation of model libraries or other things that extend the functionality of
Simphony should fit easily into the existing framework.

NOTE: THIS VERSION IS NOT BACKWARDS COMPATIBLE.

Added
- [31](https://github.com/sequoiap/simphony/issues/31) Simphony was placed
under continuous integration to ensure code health.
- [11](https://github.com/BYUCamachoLab/simphony/issues/11) Examples and
documentation were created and are hosted online at
https://simphonyphotonics.readthedocs.io/
- Nearly the entire SiEPIC EBeam PDK Compact Model Library is available.
- Parser compatible with circuit files exported by SiEPIC-Tools in KLayout
allows Simphony to perform simulations on files created in KLayout.

Changed
- Entire package was rewritten from previous versions.

Removed
- Simphony simulation engine completely rewritten. Former scripts are not
compatible with this version and will need to be rewritten.

---

Some minor updates are included in this version.

Added
- [10](https://github.com/sequoiap/simphony/issues/10) ENHANCEMENT: Extras can
now be passed in to the ebeam waveguide model to modify default parameters
for central wavelength, effective index, group index, and dispersion.

Changed
- README updated

Removed
-

---

[0.2.0](https://github.com/BYUCamachoLab/simphony/tree/v0.2.0) - <small>2019-08-07</small>

Due to a complete code overhaul, this version is incompatible with previous
versions of Simphony. Typically, when a release is incompatible, the MAJOR
version number is bumped. However, since this software is still in the first
stages of development, bumping the MINOR version symbolizes major changes
until the version 1 is officially released.

Added
- Since we only need two of the 'connect' algorithms from scikit-rf, they were
brought over into this library to reduce the number of dependencies.
- Added test cases using the pytest framework
- Models are now cached during simulation to avoid repeating calculations
- Logging is now the method of choice for getting info messages from the program
- Simulation runtime is now logged at the INFO level
- Documentation! Built using Sphinx.

Changed
- Simphony is now segmented into three modules: core, simulation, and DeviceLibrary
- New component implementation methods (simphony.core.base.ComponentModel)
- Rebuilt device library to match new component implementations
- Device library no long reads data files for s-parameters, as they've been
converted and are now stored as .npz files.
- [6](https://github.com/sequoiap/simphony/pull/6): ENHANCEMENT: Framework
rebuild, allows for easier implementation of custom component models and
libraries.

Removed
- Dependencies on scikit-rf, jsons, and matplotlib have been removed.
- Netlist export functionality no longer exists since netlists can be
scripted in Python.

---

Added
- Waveguide parameters now calculated based off new regression model provided by Easton Potokar.
- [3](https://github.com/sequoiap/simphony/pull/3): ENHANCEMENT: Add functionality for multi-input simulations

Changed
- Dependencies are slightly more flexible now, not pinned to a specific version of software.

Removed
- Dependencies on TensorFlow, Keras, and h5py have been removed since the waveguide model has been replaced.
- Removed the "settings" gui, as it's been migrated over to SiEPIC-Simphony.
- Persistent settings that should only really be in the SiEPIC-Simphony integration package were removed from here and implemented there instead.

---

Added
- N/A

Changed
- Dependencies are slightly more flexible now, not pinned to a specific version of software

Removed
- circuit_simulation.py, a GUI primarily from when package was part of SiEPIC-Tools
- monte_carlo_simulation.py, a GUI primarily from when package was part of SiEPIC-Tools

---

Added
- setup.py includes data files in build now

Changed
- N/A

Removed
- N/A