# -*- coding: utf-8 -*-
"""
绘制线性分类器与决策树分类边界的对比图。
"""
import os
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.colors import ListedColormap
# ── 1. 基本设置 ─────────────────────────────────────────────
os.makedirs("./figs", exist_ok=True)
# 图形颜色：0 表示黄色区域，1 表示绿色区域
YELLOW = "#FFF09A"
GREEN = "#97F99A"
CMAP = ListedColormap([YELLOW, GREEN])
# 坐标范围
x_min, x_max = -2.2, 2.2
y_min, y_max = -2.2, 2.2
# 用网格点生成二维平面
x1 = np.linspace(x_min, x_max, 500)
x2 = np.linspace(y_min, y_max, 500)
X1, X2 = np.meshgrid(x1, x2)
# ── 2. 构造真实的分类区域 ──────────────────────────────────
Z_linear = (X2 >= 0.55 * X1 - 0.50).astype(int)
Z_nonlinear = np.where((X1 >= -1.0) & (X2 <= 1.0), 0, 1)
# ── 3. 辅助函数 ────────────────────────────────────────────
def format_axis(ax):
"""统一设置坐标轴格式。"""
ax.set_xlim(x_min, x_max)
ax.set_ylim(y_min, y_max)
ax.set_aspect("equal", adjustable="box")
ax.set_xlabel(r"$X_1$", fontsize=11)
ax.set_ylabel(r"$X_2$", fontsize=11)
ax.set_xticks([-2, -1, 0, 1, 2])
ax.set_yticks([-2, -1, 0, 1, 2])
ax.tick_params(axis="both", labelsize=9, direction="out")
for spine in ax.spines.values():
spine.set_linewidth(0.8)
def draw_region(ax, Z):
"""绘制分类区域。"""
ax.contourf(X1, X2, Z, levels=[-0.5, 0.5, 1.5], cmap=CMAP, alpha=1.0)
# ── 4. 绘制图形 ────────────────────────────────────────────
fig, axes = plt.subplots(2, 2, figsize=(7.2, 7.0))
xx = np.linspace(x_min, x_max, 200)
# 左上：线性边界 + 线性模型
axes[0, 0].contourf(X1, X2, Z_linear, levels=[-0.5, 0.5, 1.5], cmap=CMAP, alpha=1.0)
axes[0, 0].plot(xx, 0.55 * xx - 0.50, color="black", linewidth=1.8)
format_axis(axes[0, 0])
# 右上：线性边界 + 决策树模型
axes[0, 1].contourf(X1, X2, Z_linear, levels=[-0.5, 0.5, 1.5], cmap=CMAP, alpha=1.0)
axes[0, 1].hlines(y=-0.40, xmin=x_min, xmax=x_max, color="black", linewidth=1.8)
axes[0, 1].vlines(x=-0.90, ymin=y_min, ymax=-0.40, color="black", linewidth=1.8)
axes[0, 1].hlines(y=-1.30, xmin=x_min, xmax=-0.90, color="black", linewidth=1.8)
axes[0, 1].hlines(y=0.30, xmin=x_min, xmax=1.00, color="black", linewidth=1.8)
axes[0, 1].vlines(x=1.00, ymin=-0.40, ymax=0.30, color="black", linewidth=1.8)
format_axis(axes[0, 1])
# 左下：非线性边界 + 线性模型
axes[1, 0].contourf(X1, X2, Z_nonlinear, levels=[-0.5, 0.5, 1.5], cmap=CMAP, alpha=1.0)
axes[1, 0].plot(xx, xx, color="black", linewidth=1.8)
format_axis(axes[1, 0])
# 右下：非线性边界 + 决策树模型
axes[1, 1].contourf(X1, X2, Z_nonlinear, levels=[-0.5, 0.5, 1.5], cmap=CMAP, alpha=1.0)
axes[1, 1].vlines(x=-1.00, ymin=y_min, ymax=y_max, color="black", linewidth=1.8)
axes[1, 1].hlines(y=1.00, xmin=-1.00, xmax=x_max, color="black", linewidth=1.8)
format_axis(axes[1, 1])
plt.subplots_adjust(wspace=0.28, hspace=0.35)
fig.savefig("./figs/ml_tree_fig02_linear_tree_boundary.png", dpi=300, bbox_inches="tight")
fig.savefig("./figs/ml_tree_fig02_linear_tree_boundary.svg", bbox_inches="tight")
plt.show()
