import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.patches as patches

plt.rcParams["font.family"] = "Times New Roman"
plt.rcParams["font.size"] = 10

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# 1. 构造数据（行和列完全按照你给的表格顺序）

# cols = [
#     "LA19", "LA21", "itw", "CFAD_seen", "CFAD_unseen",
#     "ST_codecfake", "omni-vocder", "omni-codec", "omni-adult",
#     "omni-child", "omni-elder", "omni-lang_zh", "omni-lang_en",
#     "omni-test", "omni-attack"
# ]

# # 数据矩阵（11行 × 15列），与你给出的表格完全一致
# data = np.array([
#     [ 0.16,  0.78, 10.41, 14.74, 22.53, 40.71, 34.77, 38.62, 30.34, 38.12, 39.17, 36.20, 38.05, 36.70, 44.87],
#     [ 2.77,  6.91, 10.97,  0.13, 24.87, 12.32, 21.16, 26.72, 19.72, 24.98, 27.30, 21.50, 26.70, 24.09, 42.62],
#     [ 2.55,  7.02,  5.53,  8.38, 24.29,  0.01, 26.17, 13.67, 16.58, 23.23, 20.33, 18.10, 19.67, 19.95, 38.20],
#     [ 4.73, 12.38,  2.54,  0.76, 26.91, 17.68,  0.68, 18.25, 11.58,  8.90, 12.43, 12.43, 11.59, 14.11, 10.97],
#     [ 5.85, 10.86,  2.12,  4.28, 24.61,  0.64, 16.27,  0.27, 10.52, 10.13,  9.83,  9.83, 10.58, 15.75, 10.41],
#     [ 1.80, 10.49,  2.34,  0.71, 24.42,  1.43, 11.85, 11.23,  0.45, 26.69,  7.48,  6.15, 17.90, 14.44, 11.56],
#     [ 2.27,  8.45,  2.06,  1.00, 24.31,  2.99,  3.42,  2.25,  2.82,  0.76,  4.88,  2.81,  2.81, 10.93,  2.86],
#     [ 1.40,  4.43,  1.94,  0.25, 24.43,  1.66, 10.71, 10.01,  3.70, 24.79,  1.45,  4.08, 15.35, 11.63, 10.36],
#     [ 1.93,  9.27,  2.34,  0.11, 23.80,  1.35,  7.66,  6.89,  1.72, 13.41,  4.29,  1.02, 11.43, 11.29,  7.32],
#     [ 1.05,  5.00,  1.76,  0.33, 22.89,  1.01,  3.70,  2.50,  1.16,  3.64,  3.93,  4.81,  0.94, 11.00,  3.14],
#     [ 1.68,  5.99,  1.68,  0.11, 23.59,  0.55,  1.03,  0.56,  0.50,  0.68,  1.31,  0.96,  0.69,  0.82,  8.01]
# ])

cols = [
    "LA19", "LA21", "itw", "CFAD", 
    "ST_codecfake", "omni-vocder", "omni-codec", "omni-adult",
    "omni-child", "omni-elder", "omni-lang_zh", "omni-lang_en",
    "omni-test", "omni-attack"
]
rows = [
    "LA19", "CFAD", "ST_codecfake", "omni-vocder", "omni-codec",
    "omni-adult", "omni-child", "omni-elder", "omni-lang_zh",
    "omni-lang_en", "OmniSynSpoof"
]
# 数据矩阵（11行 × 15列），与你给出的表格完全一致
data = np.array([
    [ 0.16,  0.78, 10.41, 14.74,  40.71, 34.77, 38.62, 30.34, 38.12, 39.17, 36.20, 38.05, 36.70, 44.87],
    [ 2.77,  6.91, 10.97,  0.13,  12.32, 21.16, 26.72, 19.72, 24.98, 27.30, 21.50, 26.70, 24.09, 42.62],
    [ 2.55,  7.02,  5.53,  8.38,   0.01, 26.17, 13.67, 16.58, 23.23, 20.33, 18.10, 19.67, 19.95, 38.20],
    [ 4.73, 12.38,  2.54,  0.76,  17.68,  0.68, 18.25, 11.58,  8.90, 12.43, 12.43, 11.59, 10.97,14.11, ],
    [ 5.85, 10.86,  2.12,  4.28,   0.64, 16.27,  0.27, 10.52, 10.13,  9.83,  9.83, 10.58, 10.41,15.75, ],
    [ 1.80, 10.49,  2.34,  0.71,  1.43, 11.85, 11.23,  0.45, 26.69,  7.48,  6.15, 17.90, 11.56,14.44, ],
    [ 2.27,  8.45,  2.06,  1.00,   2.99,  3.42,  2.25,  2.82,  0.76,  4.88,  2.81,  2.81, 2.86,10.93,  ],
    [ 1.40,  4.43,  1.94,  0.25,   1.66, 10.71, 10.01,  3.70, 24.79,  1.45,  4.08, 15.35, 10.36,11.63, ],
    [ 1.93,  9.27,  2.34,  0.11,   1.35,  7.66,  6.89,  1.72, 13.41,  4.29,  1.02, 11.43, 7.32,11.29,  ],
    [ 1.05,  5.00,  1.76,  0.33,   1.01,  3.70,  2.50,  1.16,  3.64,  3.93,  4.81,  0.94, 3.14,11.00,  ],
    [ 1.68,  5.99,  1.68,  0.11,   0.55,  1.03,  0.56,  0.50,  0.68,  1.31,  0.96,  0.69,  0.82,  8.01]
])

df = pd.DataFrame(data, index=rows, columns=cols)

# 2. 设置字体（避免 Times New Roman 缺失警告）
plt.rcParams['font.family'] = 'sans-serif'
plt.rcParams['font.sans-serif'] = ['DejaVu Sans', 'Arial', 'Helvetica']

# 3. 绘图
plt.figure(figsize=(18, 8))
# 颜色映射：浅色=低EER（好），深色=高EER（差），并限制颜色范围
ax = sns.heatmap(df, annot=True, fmt=".2f", cmap="Blues_r", linewidths=0.5,
                 cbar_kws={'label': 'EER (%)'}, vmin=0, vmax=50)

# 4. 高亮行列名称相同的单元格（自身测试）
for i in range(len(rows)):
    for j in range(len(cols)):
        if rows[i] == cols[j]:
            ax.add_patch(plt.Rectangle((j, i), 1, 1, fill=False, edgecolor='blue', lw=2))
        elif rows[i] =='OmniSynSpoof' and cols[j]=='omni-test':
            ax.add_patch(plt.Rectangle((j, i), 1, 1, fill=False, edgecolor='blue', lw=2))

# 5. 标题和标签
# plt.title("EER Heatmap", fontsize=14, fontweight='bold')
plt.xlabel("Test Dataset", fontsize=12)
plt.ylabel("Train Dataset", fontsize=12)
plt.xticks(rotation=45, ha='right')
plt.yticks(rotation=0)
plt.tight_layout()
# plt.savefig("eer_heatmap.png", dpi=150)
# plt.show()


plt.savefig("heatmap_usenix.pdf",dpi=600,bbox_inches="tight")
plt.savefig("heatmap_usenix.png",dpi=600,bbox_inches="tight")

plt.show()