Sympy-plot-backends

What's Changed

* Added new plotting functions:

* ``plot3d_revolution`` to create surface of revolution.

* ``plot_parametric_region``, still in development.

* ``MatplotlibBackend``:

* Fixed bug with colormaps and normalization.

* Improved update speed when dealing with parametric domain coloring plots.

* Improved ``zlim`` support on ``K3DBackend`` for interactive widget plots.

* Fixed bug with parametric interactive widget plots and ``PlotlyBackend``: the update speed is now decent.

* Series:

* Moved ``LineOver1DRangeSeries._detect_poles`` to ``_detect_poles_helper``.

* ``plot_complex`` and ``plot_real_imag``: the input expression is no longer wrapped by symbolic ``re()`` or ``im()``. Instead, the necessary processing is done on the series after the complex function has been evaluated. This improves performance.

* ``Parametric2DLineSeries`` now support ``detect_poles``.

* Implemented support for ``color_func`` keyword argument on ``plot_list`` and ``plot_complex_list``.

* Added ``extras_require`` to ``setup.py``:

* by default, ``pip install sympy_plot_backends`` will install only the necessary requirements to get non-interactive plotting to work with Matplotlib.
* use ``pip install sympy_plot_backends[all]`` to install all other packages: panel, bokeh, plotly, k3d, vtk, ...

* Documentation:

* Improved examples.

* Added examples with ``PlotlyBackend``.

What's Changed

* Implemented the ``plot3d_spherical`` function to plot functions in spherical coordinates.

* Added the ``wireframe`` option to ``plot3d``, ``plot3d_parametric_surface`` and ``plot3d_spherical`` to add grid lines over the surface.

* Fixed bug with ``plot3d`` and ``plot_contour`` when dealing with instances of ``BaseScalar``.

* Added ``normalize`` keyword argument to ``plot_vector`` and ``plot_complex_vector`` to visualize quivers with unit length.

* Improve documentation of ``plot_vector`` and ``plot_complex_vector``.

* Improved test coverage on complex and vector plotting functions.

* Improvements on ``PlotlyBackend``:

* it is now be able to plot more than 14 2d/3d parametric lines when ``use_cm=False``.

* improved logic to show colorbars on 3D surface plots.

* added support for custom aspect ratio on 3D plots.

* Improved support for ``xlim``, ``ylim``, ``zlim`` on ``K3DBackend``.

* Series:

* Fixed bug with uniform evaluation while plotting numerical functions.

* Fixed bug with ``color_func``.

* Added transformation keyword arguments ``tx, ty, tz`` to parametric series.

* Breaks:

* Inside ``plot_parametric`` and ``plot3d_parametric_line``, the ``tz`` keyword argument has been renamed to ``tp``.

* Removed Mayavi from setup dependencies. Mayavi is difficult to install: can't afford the time it requires for proper setup and testing. ``MayaviBackend`` is still available to be used "as is".

What's Changed

* Reintroduced ``MayaviBackend`` to plot 3D symbolic expressions with Mayavi. Note that interactive widgets are still not supported by this backend.

* ``plot_contour`` is now able to create filled contours or line contours on backends that supports such distinction. Set the ``is_filled`` keyword argument to choose the behaviour.

* Implemented interactive widget support for ``plot_list``.

* Implemented back-compatibility-related features with SymPy.

* Fixed bugs with ``PlaneSeries``:

* Data generation for vertical planes is now fixed.
* ``K3DBackend`` is now able to plot this series.
* Similar to other 3D surfaces, planes will be plotted with a solid color.

* Fixed bug with ``Vector3DSeries``: the discretized volume is now created with Numpy's ``meshgrid`` with ``indexing='ij'``. This improves the generation of 3D streamlines.

* Fixed bug with ``plot3d`` and ``plot_contour``: when ``params`` is provided the specified backend will be instantiated.

* Fixed bug with ``K3DBackend`` and ``plot3d_implicit``.

What's Changed

Fixed bad character for conda builds.

For a full list of changes please look at: [Release Notes](https://github.com/Davide-sd/sympy-plot-backends/releases/tag/v1.4.0).

What's Changed

* Added support for plotting numerical vectorized functions. Many of the plotting functions exposed by this module are now able to deal with both symbolic expressions as well as numerical functions. This extends the scope of this module, as it is possible to use it directly with numpy and lambda functions. For example, the following is now supported:

import numpy as np
plot(lambda t: np.cos(x) * np.exp(-x / 5), ("t", 0, 10))

* Added support for vector from the ``sympy.physics.mechanics`` module in the ``plot_vector`` function.

* Implemented keyword argument validator: if a user writes a misspelled keyword arguments, a warning message will be raised showing one possible alternative.

What's Changed

* Added ``used_by_default`` inside default options for adaptive algorithm. This let the user decide wheter to use adaptive algorithm or uniform meshing by default for line plots.

* Fix the axis labels for the ``plot_complex_vector`` function.

* Improved a few examples in the docstring of ``plot_vector`` and ``plot_complex_vector``.

* Fixed bug with interactive update of ``plot_vector`` inside ``MatplotlibBackend``.

* Improvements to the code in preparation for merging this module into Sympy:

* Small refactoring about the label generation: previously, the string and latex representations were generated at different times and in different functions. Now, they are generated simultaneously inside the ``__init__`` method of a data series.

* Changes in names of functions that are meant to remain private:

* ``adaptive_eval`` -> ``_adaptive_eval``.
* ``_uniform_eval`` -> ``_uniform_eval_helper``
* ``uniform_eval`` -> ``_uniform_eval``
* ``_correct_size`` -> ``_correct_shape``
* ``get_points`` -> ``_get_points``

v1.2.0.d
Updated requirements for conda builds.

For a full list of changes please look at: [Release Notes](https://github.com/Davide-sd/sympy-plot-backends/releases/tag/v1.2.0).