What's Changed
If you are upgrading from a previous version, you should run the following code to load the new configuration settings:
py
from spb.defaults import reset
reset()
* Breaking changes:
* Refactoring of ``*Series`` classes. All ``*InteractiveSeries`` classes have been removed. The interactive functionalities have been integrated on regular ``*Series``. This greatly simplifies the code base, meaning bug fixes should take less time to implement.
* Refactoring of ``iplot`` to take into account the aforementioned changes. In particular, interactive widget plots are now tighly integrated into the usual plotting functions. This improves user experience and simplifies the code base.
* The ``spb.interactive.create_series`` function has been removed.
* Changed the default evaluation algorithm to a uniform sampling strategy, instead of the adaptive algorithm. The latter is still available, just set ``adaptive=True`` on the plotting functions that support it. The motivation behind this change is that the adaptive algorithm is usually much slower to produce comparable results: by default, the uniform sampling strategy uses 1000 discretization points over the specified range (users can increase it or decrease it), which is usually enough to smoothly capture the function.
It also simplifies the dependencies of the module: now, the adaptive algorithm is not required by the plotting module to successfully visualize symbolic expressions, hence it is not installed. If users need the adaptive algorithm, they'll have to follow the [adaptive module installation instructions](https://github.com/python-adaptive/adaptive).
* Improved support for plotting summations.
* Implemented wireframe lines for 3D complex plots.
* Interactive widget plots.
* Users can now chose the interactive module to be used:
* ``ipywidgets``: new in this release. It is the default one.
* ``panel``: the same, old one.
Please, read the documentation about the interactive sub-module to learn more about them, and how to chose one or the other.
* Implemented the ``template`` keyword argument for interactive widget plots with Holoviz's Panel and ``servable=True``: user can further customize the layout of the web application, or can provide their own Panel's templates.
* The module is now fully interactive. Thanks to the ``prange`` class, it is possible to specify parametric ranges. Explore the examples in the module documentation to find out how to use it.
* ``color_func`` now support symbolic expressions.
* ``line_color`` and ``surface_color`` are now deprecated in favor of ``color_func``.
* ``plot_implicit``:
* now it supports interactive-widget plots, when ``adaptive=False``.
* not it support ``rendering_kw`` for plots created with ``adaptive=True``.
* improved logic dealing with legends. When plotting multiple regions, rectangles will be visible on the legend. When plotting multiple lines, lines will be visible on the legend.
* Removed ``tutorials`` folder containing Jupyter notebooks. The documentation contains plently of examples: the notebooks were just reduntant and difficult to maintain.
* ``MatplotlibBackend``: implemented support for ``ipywidgets``.
* ``PlotlyBackend``:
* fixed bug with interactive update of lines.
* implemented support for ``ipywidgets``.
* ``BokehBackend``:
* improved support for Bokeh 3.0.
* removed ``update_event`` because it became a redundant feature now that the module is fully parametric.
* ``plot_contour``: added the ``clabels`` keyword argument to show/hide contour labels.
* Documentation is now able to show interactive widget plots with K3D-Jupyter.
* conda package is now built and made available through the conda-forge channel. This greatly simplify the workflow and should allow an easier installation with conda.