"""
Lollipop comparisons
====================

Vertical and horizontal lollipop charts in a publication layout.

Why UltraPlot here?
-------------------
UltraPlot adds lollipop plot methods that mirror bar plotting while exposing
simple styling for stems and markers. This plot type is not built into
Matplotlib.

Key functions: :py:meth:`ultraplot.axes.PlotAxes.lollipop`, :py:meth:`ultraplot.axes.PlotAxes.lollipoph`.

See also
--------
* :doc:`1D plot types </1dplots>`
"""

import numpy as np
import pandas as pd

import ultraplot as uplt

rng = np.random.default_rng(11)
categories = ["Alpha", "Beta", "Gamma", "Delta", "Epsilon", "Zeta"]
values = np.sort(rng.uniform(0.4, 1.3, len(categories)))
data = pd.Series(values, index=categories, name="score")

fig, axs = uplt.subplots(ncols=2, share=0, refwidth=2.8)
axs[0].lollipop(data, stemcolor="black", marker="o", color="C0")
axs[0].format(title="Vertical lollipop", xlabel="Category", ylabel="Score")

axs[1].lollipoph(data, stemcolor="black", marker="o", color="C1")
axs[1].format(title="Horizontal lollipop", xlabel="Score", ylabel="Category")

axs.format(abc=True, abcloc="ul", suptitle="Lollipop charts for ranked metrics")

fig.show()