"""
Diverging colormap
==================

Use a diverging colormap with centered normalization.

Why UltraPlot here?
-------------------
UltraPlot can automatically detect diverging datasets (spanning negative and
positive values) and apply a diverging colormap with a centered normalizer.
This ensures the "zero" point is always at the center of the colormap.

Key functions: :py:class:`ultraplot.colors.DivergingNorm`, :py:meth:`ultraplot.axes.PlotAxes.pcolormesh`.

See also
--------
* :doc:`Colormaps </colormaps>`
* :doc:`Normalizers </colors>`
"""

import numpy as np

import ultraplot as uplt

# Generate data with negative and positive values
x = np.linspace(-5, 5, 100)
y = np.linspace(-5, 5, 100)
X, Y = np.meshgrid(x, y)
Z = np.sin(X) * np.cos(Y) + 0.5 * np.cos(X * 2)

fig, axs = uplt.subplots(ncols=2, refwidth=3)

# 1. Automatic diverging
# UltraPlot detects Z spans -1 to +1 and uses the default diverging map
m1 = axs[0].pcolormesh(X, Y, Z, cmap="Div", colorbar="b", center_levels=True)
axs[0].format(title="Automatic diverging", xlabel="x", ylabel="y")

# 2. Manual control
# Use a specific diverging map and center it at a custom value
m2 = axs[1].pcolormesh(
    X,
    Y,
    Z + 0.5,
    cmap="ColdHot",
    diverging=True,
    colorbar="b",
    center_levels=True,
)
axs[1].format(title="Manual center at 0.5", xlabel="x", ylabel="y")

axs.format(suptitle="Diverging colormaps and normalizers")
fig.show()