"""
Configuration and Styling
==========================

This example demonstrates how to configure PubliPlots using rcParams,
set global styles, and customize various plotting parameters.
"""

import publiplots as pp
import pandas as pd
import numpy as np

# %%
# Understanding rcParams
# ----------------------
# PubliPlots uses an rcParams system similar to matplotlib and seaborn.
# You can access both matplotlib parameters and PubliPlots-specific parameters
# through the same unified interface.

# View some default PubliPlots parameters
print("PubliPlots Custom Parameters:")
print(f"  Default color: {pp.rcParams['color']}")
print(f"  Default alpha: {pp.rcParams['alpha']}")
print(f"  Default edgecolor: {pp.rcParams['edgecolor']}")
print(f"  Default capsize: {pp.rcParams['capsize']}")
print(f"  Hatch mode: {pp.rcParams['hatch_mode']}")

print("\nMatplotlib Parameters (via pp.rcParams):")
print(f"  Axes size (mm): {pp.rcParams['subplots.axes_size']}")
print(f"  Line width: {pp.rcParams['lines.linewidth']}")
print(f"  Font size: {pp.rcParams['font.size']}")
print(f"  DPI: {pp.rcParams['savefig.dpi']}")

# %%
# Publication-Grade Defaults
# ---------------------------
# PubliPlots applies publication-grade styling automatically on import.
# No style function calls are needed — you get print-ready figures by default.

# Create sample data for demonstration
np.random.seed(100)
sample_data = pd.DataFrame({
    'category': np.repeat(['A', 'B', 'C'], 10),
    'value': np.concatenate([
        np.random.normal(50, 8, 10),
        np.random.normal(70, 10, 10),
        np.random.normal(85, 12, 10),
    ])
})

print("\nDefault PubliPlots Style (applied on import):")
print(f"  Axes size (mm): {pp.rcParams['subplots.axes_size']}")
print(f"  Font size: {pp.rcParams['font.size']}")
print(f"  DPI: {pp.rcParams['savefig.dpi']}")

ax = pp.barplot(
    data=sample_data,
    x='category',
    y='value',
    title='Default Publication-Grade Style',
    errorbar='se',
    palette='pastel'
)
pp.show()

# %%
# Customizing Individual Parameters
# ----------------------------------
# You can customize any parameter by setting it through pp.rcParams.

# Set custom defaults
pp.rcParams['color'] = '#E67E7E'  # Change default color to red
pp.rcParams['alpha'] = 0.3        # Increase default transparency
pp.rcParams['capsize'] = 0.15     # Larger error bar caps
pp.rcParams['hatch_mode'] = 2     # Medium density hatch patterns

# Also customize matplotlib parameters
pp.rcParams['subplots.axes_size'] = (80, 50)  # Wider axes (mm)
pp.rcParams['lines.linewidth'] = 2.5           # Thicker lines
pp.rcParams['font.size'] = 11                  # Slightly larger font

# Create plot with custom defaults
ax = pp.barplot(
    data=sample_data,
    x='category',
    y='value',
    title='Plot with Custom rcParams',
    errorbar='se',
)
pp.show()

# Reset to default (reverts to matplotlib defaults)
pp.reset_style()

# %%
# Using Color Palettes
# --------------------
# PubliPlots provides easy access to color palettes.

# Create data for palette demonstration
palette_data = pd.DataFrame({
    'group': np.repeat(['Group A', 'Group B', 'Group C', 'Group D'], 12),
    'measurement': np.concatenate([
        np.random.normal(45, 7, 12),
        np.random.normal(60, 8, 12),
        np.random.normal(75, 9, 12),
        np.random.normal(90, 10, 12),
    ])
})

# Using built-in palettes
fig, axes = pp.subplots(2, 2, axes_size=(85, 70))

# Pastel palette
pp.barplot(
    data=palette_data,
    x='group',
    y='measurement',
    hue='group',
    palette='pastel',
    errorbar='se',
    title='Pastel Palette',
    ax=axes[0, 0]
)

# Tab10 palette
pp.barplot(
    data=palette_data,
    x='group',
    y='measurement',
    hue='group',
    palette='tab10',
    errorbar='se',
    title='Tab10 Palette',
    ax=axes[0, 1]
)

# Custom color list
custom_colors = ['#E67E7E', '#75B375', '#E6B375', '#8E8EC1']
pp.barplot(
    data=palette_data,
    x='group',
    y='measurement',
    hue='group',
    palette=custom_colors,
    errorbar='se',
    title='Custom Color List',
    ax=axes[1, 0]
)

# Custom color dictionary
custom_dict = {
    'Group A': '#E67E7E',
    'Group B': '#75B375',
    'Group C': '#E6B375',
    'Group D': '#8E8EC1'
}
pp.barplot(
    data=palette_data,
    x='group',
    y='measurement',
    hue='group',
    palette=custom_dict,
    errorbar='se',
    title='Custom Color Dictionary',
    ax=axes[1, 1]
)

pp.show()

# %%
# Context-Based Styling
# ----------------------
# Temporarily override parameters for specific plots without
# changing global settings.

# Create sample scatter data
np.random.seed(200)
scatter_data = pd.DataFrame({
    'x': np.random.randn(80),
    'y': np.random.randn(80),
    'category': np.random.choice(['Type 1', 'Type 2', 'Type 3'], 80)
})

# Plot 1: Default settings
ax = pp.scatterplot(
    data=scatter_data,
    x='x',
    y='y',
    hue='category',
    title='Default Settings',
    palette='pastel',
    alpha=0.2,
)
pp.show()

# Plot 2: Override alpha and color for this plot only
ax = pp.scatterplot(
    data=scatter_data,
    x='x',
    y='y',
    hue='category',
    title='Custom Alpha (Higher Opacity)',
    palette='pastel',
    alpha=0.5,  # Override default alpha just for this plot
)
pp.show()

# %%
# Saving Figures with Custom Settings
# ------------------------------------
# Control output quality and format when saving figures.

# Create a sample figure
ax = pp.barplot(
    data=sample_data,
    x='category',
    y='value',
    title='Sample Figure for Saving',
    errorbar='se',
    palette='pastel'
)

# Save with different settings (uncomment to actually save)
# pp.savefig('figure_low_res.png', dpi=150)      # Lower resolution
# pp.savefig('figure_high_res.png', dpi=300)     # High resolution
# pp.savefig('figure_vector.pdf')                # Vector format (PDF)
# pp.savefig('figure_vector.svg')                # Vector format (SVG)

print("Figure saving examples (commented out to prevent file creation)")
print("  - PNG at 150 DPI (web/presentations)")
print("  - PNG at 300 DPI (publications)")
print("  - PDF (vector, editable)")
print("  - SVG (vector, web-friendly)")

pp.show()

# %%
# Best Practices Summary
# ----------------------
# 1. Publication-grade defaults are applied automatically on import
# 2. Customize global defaults with pp.rcParams for consistency
# 3. Override parameters per-plot when needed using function arguments
# 4. Use pp.reset_style() to revert to matplotlib defaults if needed
# 5. Use color palettes for consistent coloring across figures
# 6. Save in vector formats (PDF/SVG) for publications
# 7. Use hatch patterns for black-and-white publications

print("\nConfiguration complete!")

# %%
# Asymmetric Grids
# ----------------
# ``pp.subplots`` accepts ``width_ratios=`` and ``height_ratios=`` to build
# asymmetric grids. ``axes_size`` remains the per-cell budget: total grid
# width is ``axes_size[0] * ncols`` and the ratios split that budget across
# columns proportionally (rows analogous). Equal ratios recover the uniform
# case. The canonical use is a seaborn-style JointGrid: a large main panel
# flanked by thin marginal strips.

rng = np.random.default_rng(42)
n = 10_000
cluster_a = rng.multivariate_normal([-1.5, -1.0], [[1.0, 0.4], [0.4, 1.0]], n // 2)
cluster_b = rng.multivariate_normal([2.0, 1.5], [[1.2, -0.3], [-0.3, 0.8]], n // 2)
joint = pd.DataFrame(np.vstack([cluster_a, cluster_b]), columns=["x", "y"])

fig, axes = pp.subplots(
    2, 2,
    axes_size=(45, 35),
    width_ratios=[7, 2],
    height_ratios=[2, 5],
)
axes[0, 1].set_visible(False)

pp.hexbinplot(
    data=joint, x="x", y="y",
    gridsize=20,
    xlabel="x", ylabel="y",
    ax=axes[1, 0],
)
pp.histplot(
    data=joint, x="x", bins=30,
    xlabel="", ylabel="",
    ax=axes[0, 0],
)
pp.histplot(
    data=joint, y="y", bins=30,
    xlabel="", ylabel="",
    ax=axes[1, 1],
)
pp.show()