# 1. 装 Miniconda（官网下载安装包，按提示装即可）

# 2. 建独立环境
conda create -n sciviz python=3.11 -y
conda activate sciviz

# 3. 一次装齐
conda install numpy pandas scipy matplotlib seaborn scikit-learn jupyter -y
pip install plotly missingno scienceplots

# 4. 验证（在 Jupyter 里跑）
import matplotlib.pyplot as plt
plt.plot([1, 2, 3], [1, 4, 9])
plt.show()

import numpy as np

# 创建：4 种主要方式
a = np.array([1, 2, 3])              # 从 list
b = np.zeros((3, 4))                  # 全零，3×4
c = np.linspace(0, 2*np.pi, 100)      # 0 到 2π 之间 100 个点（画图最常用）
d = np.arange(0, 10, 2)               # 等差：[0,2,4,6,8]

# 索引：取行 / 取列 / 按条件
m = np.arange(20).reshape(4, 5)
m[1, 2]         # 单元素
m[:, 2]         # 第 2 列
m[m > 10]       # 按条件筛选

# 广播：不同形状自动对齐运算
m + np.array([1, 2, 3, 4, 5])   # (4,5) + (5,) → 每行都加这个向量

import numpy as np

a = np.array([[1, 2, 3],
              [4, 5, 6]])    # shape: (2, 3)

# 全局聚合
a.mean()           # 3.5
a.std()            # 1.71

# 沿轴聚合（重点）
a.mean(axis=0)     # [2.5, 3.5, 4.5]  消掉第 0 维 → 每列均值
a.mean(axis=1)     # [2.0, 5.0]       消掉第 1 维 → 每行均值

# NaN 处理：用 nan 系列函数
b = np.array([1, 2, np.nan, 4])
np.nanmean(b)      # 2.33（忽略 NaN）

# 可复现的随机数（新 API）
rng = np.random.default_rng(seed=42)
rng.normal(0, 1, size=100)      # 100 个标准正态
rng.uniform(0, 10, size=(3, 4)) # 3×4 均匀分布

import pandas as pd

# 1. 读取
df = pd.read_csv('data.csv', parse_dates=['date'], encoding='utf-8')

# 2. 看数据（三句魔咒）
df.head()        # 前 5 行
df.info()        # 列类型、非空数
df.describe()    # 数值列的统计摘要

# 3. 清洗
df.isna().sum()                              # 每列缺失数
df = df.dropna(subset=['important_col'])     # 删关键列缺失
df['x'] = df['x'].fillna(df['x'].median())   # 中位数填充

# 4. 选行选列
df[df['age'] > 25]                           # 条件筛选
df.loc[df['age'] > 25, ['name', 'income']]   # 同时筛行选列

# 5. 分组聚合
df.groupby('city')['income'].mean()                              # 按城市算均收入
df.groupby('city').agg({'income': 'mean', 'age': 'median'})      # 多列多函数

# 6. 时间序列
df = df.set_index('date')
df.resample('M').mean()                       # 月均
df['ma7'] = df['value'].rolling(7).mean()     # 7 日滚动均

import matplotlib.pyplot as plt
import numpy as np
x = np.linspace(0, 10, 50)

# 方式 1：pyplot 风格（脚本式，省事）
plt.plot(x, np.sin(x))
plt.title("Sin")
plt.show()

# 方式 2：面向对象（推荐）
fig, ax = plt.subplots(figsize=(7, 4))
ax.plot(x, np.sin(x))
ax.set_title("Sin")
plt.show()

import matplotlib.pyplot as plt

# 一张图一个坐标系
fig, ax = plt.subplots(figsize=(7, 4))

# 一张图四个坐标系（2 行 2 列）
fig, axes = plt.subplots(2, 2, figsize=(10, 8))
axes[0, 0].plot([1, 2, 3])        # 左上
axes[0, 1].scatter([1, 2], [3, 4]) # 右上
axes[1, 0].bar(['A', 'B'], [3, 5]) # 左下
axes[1, 1].hist([1, 2, 2, 3, 3])   # 右下

plt.tight_layout()  # 自动调整子图间距
plt.show()

fig, ax = plt.subplots(figsize=(6, 4))
ax.plot([1, 2, 3], [1, 4, 9])

# 推荐：双格式同时保存
fig.savefig('out.pdf', bbox_inches='tight')             # 矢量，投稿用
fig.savefig('out.png', dpi=300, bbox_inches='tight')    # 位图，预览/网页用

# 透明背景（嵌入 PPT 时用）
fig.savefig('out.png', dpi=300, transparent=True, bbox_inches='tight')

import matplotlib.pyplot as plt
import numpy as np
x = np.linspace(0, 2*np.pi, 50)

fig, ax = plt.subplots(figsize=(7, 4))
ax.plot(x, np.sin(x),  label='sin',  color='#2E86AB', linestyle='-',  marker='o', markevery=5)
ax.plot(x, np.cos(x),  label='cos',  color='#E63946', linestyle='--', marker='s', markevery=5)
ax.plot(x, np.sin(2*x), label='sin2x', color='#06A77D', linestyle=':',  marker='^', markevery=5)

ax.set_xlabel('x'); ax.set_ylabel('y'); ax.legend()
plt.show()

import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(42)
x = rng.normal(0, 1, 100)
y = x * 0.5 + rng.normal(0, 0.5, 100)
sizes = rng.uniform(20, 200, 100)
colors = rng.uniform(0, 1, 100)

fig, ax = plt.subplots(figsize=(7, 5))
sc = ax.scatter(x, y, s=sizes, c=colors, cmap='viridis',
                alpha=0.6, edgecolors='white', linewidth=0.5)
fig.colorbar(sc, ax=ax, label='Third Variable')
ax.set_xlabel('X'); ax.set_ylabel('Y')
plt.show()

import matplotlib.pyplot as plt
import numpy as np

categories = ['A', 'B', 'C', 'D']
control = [23, 45, 56, 78]
treatment = [25, 50, 60, 85]

x = np.arange(len(categories))
width = 0.35

fig, ax = plt.subplots(figsize=(7, 4))
ax.bar(x - width/2, control,   width, label='Control',   color='#2E86AB')
ax.bar(x + width/2, treatment, width, label='Treatment', color='#E63946')

ax.set_xticks(x)
ax.set_xticklabels(categories)
ax.set_ylabel('Value'); ax.legend()
plt.show()

import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(42)
data = np.concatenate([rng.normal(0, 1, 1000),
                        rng.normal(5, 1.5, 500)])  # 双峰分布

fig, ax = plt.subplots(figsize=(7, 4))
ax.hist(data, bins=40, color='#2E86AB', edgecolor='white', alpha=0.8)
ax.set_xlabel('Value'); ax.set_ylabel('Count')
plt.show()

import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(42)
data = [rng.normal(loc=m, scale=1, size=100) for m in [0, 1, 2, 3]]

fig, ax = plt.subplots(figsize=(7, 4))
bp = ax.boxplot(data, labels=['A', 'B', 'C', 'D'],
                patch_artist=True,           # 允许填色
                boxprops=dict(facecolor='#2E86AB', alpha=0.7),
                medianprops=dict(color='red', linewidth=2))
ax.set_ylabel('Value')
plt.show()

import matplotlib.pyplot as plt
import numpy as np
x = np.linspace(0, 2*np.pi, 100)

fig, ax = plt.subplots(figsize=(7, 4))
ax.plot(x, np.sin(x), label=r'$y = \sin(x)$')
ax.plot(x, np.cos(x), label=r'$y = \cos(x)$')

# 标签与标题
ax.set_xlabel('Angle (rad)', fontsize=12)
ax.set_ylabel('Amplitude',   fontsize=12)
ax.set_title(r'Trigonometric Functions: $\sin$ and $\cos$', fontsize=13)

# 坐标轴范围与刻度
ax.set_xlim(0, 2*np.pi)
ax.set_xticks([0, np.pi/2, np.pi, 3*np.pi/2, 2*np.pi])
ax.set_xticklabels(['0', r'$\pi/2$', r'$\pi$', r'$3\pi/2$', r'$2\pi$'])

# 图例与网格
ax.legend(loc='upper right', fontsize=10)
ax.grid(True, alpha=0.3)

plt.show()

import matplotlib.pyplot as plt
import numpy as np

# 指定颜色的 3 种方式
ax.plot(x, y, color='red')         # 命名
ax.plot(x, y, color='#2E86AB')     # HEX
ax.plot(x, y, color=(0.2, 0.5, 0.8)) # RGB 元组（0-1）

# 三类 Colormap
data = np.random.rand(10, 10)
fig, axes = plt.subplots(1, 3, figsize=(12, 3.5))
axes[0].imshow(data, cmap='viridis')      # 顺序型：从小到大
axes[1].imshow(data*2-1, cmap='RdBu_r')   # 发散型：以 0 为中心
axes[2].imshow(np.random.randint(0, 5, (10,10)), cmap='tab10')  # 定性型：离散类别
plt.show()

# 色盲友好色板（5 色版）
colors = ['#2E86AB',   # 蓝
          '#E63946',   # 红
          '#06A77D',   # 绿
          '#F4A261',   # 橙
          '#6C757D']   # 灰

# 用 seaborn 现成的色盲友好色板
import seaborn as sns
colors = sns.color_palette('colorblind')   # 10 色色盲友好

# 检查黑白可读：保存灰度图看一眼
import matplotlib.pyplot as plt
fig.savefig('test_gray.png', dpi=150)
# 用图像查看器看 → 几条线能不能区分？

import matplotlib.pyplot as plt
import numpy as np

x = np.linspace(0, 10, 50)
mean = np.sin(x)
std = 0.2

fig, ax = plt.subplots(figsize=(7, 4))

# 主曲线
ax.plot(x, mean, color='#2E86AB', linewidth=2.0,
        linestyle='-', marker='o', markevery=5, markersize=6,
        label='Mean')

# 置信区间（最常用的 fill_between）
ax.fill_between(x, mean - std, mean + std,
                color='#2E86AB', alpha=0.2, label='±1 std')

ax.legend(); plt.show()

import matplotlib.pyplot as plt

# Windows
plt.rcParams['font.sans-serif'] = ['SimHei']        # 黑体

# Mac
plt.rcParams['font.sans-serif'] = ['Arial Unicode MS']

# Linux（先安装：sudo apt install fonts-wqy-zenhei）
plt.rcParams['font.sans-serif'] = ['WenQuanYi Zen Hei']

# 通用：让 minus 正常显示
plt.rcParams['axes.unicode_minus'] = False

# 验证
fig, ax = plt.subplots()
ax.plot([1, 2, 3], [-1, -2, -3])
ax.set_title('中文标题 测试 -1')
plt.show()

import matplotlib.pyplot as plt

# 方式 1：手动设全局
plt.rcParams['figure.figsize'] = (7, 4.5)
plt.rcParams['axes.labelsize'] = 12
plt.rcParams['axes.titlesize'] = 13
plt.rcParams['lines.linewidth'] = 2.0
plt.rcParams['axes.grid'] = True
plt.rcParams['grid.alpha'] = 0.3

# 方式 2：用预设样式（瞬间切换风格）
plt.style.use('seaborn-v0_8-whitegrid')   # 推荐：清爽
# plt.style.use('ggplot')                   # 备选：R 风格
# plt.style.use('default')                  # 还原默认

# 查看所有可用样式
print(plt.style.available)

import matplotlib.pyplot as plt

# 简单：2×2 等大网格
fig, axes = plt.subplots(2, 2, figsize=(10, 7))
axes[0, 0].plot([1, 2, 3])
axes[1, 1].scatter([1, 2], [3, 4])
plt.tight_layout()

# 进阶：GridSpec 不等大网格
from matplotlib.gridspec import GridSpec
fig = plt.figure(figsize=(10, 6))
gs = GridSpec(2, 3, figure=fig)

ax1 = fig.add_subplot(gs[0, :])      # 第 0 行跨 3 列（大图）
ax2 = fig.add_subplot(gs[1, 0])      # 左下
ax3 = fig.add_subplot(gs[1, 1])      # 中下
ax4 = fig.add_subplot(gs[1, 2])      # 右下

ax1.set_title('a) Big plot on top')
ax2.set_title('b)')
ax3.set_title('c)')
ax4.set_title('d)')
plt.tight_layout()
plt.show()

import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(42)
x = np.arange(1, 11)
y = np.log(x)
yerr = 0.1 + 0.05 * x   # 误差随 x 增大

fig, ax = plt.subplots(figsize=(7, 4))
ax.errorbar(x, y, yerr=yerr,
            fmt='o-',                # 数据点 + 连线
            color='#2E86AB',
            ecolor='gray',           # 误差棒颜色
            elinewidth=1.5,
            capsize=4,               # 误差棒"帽子"长度
            capthick=1.5,
            label='Experiment')
ax.set_xlabel('x'); ax.set_ylabel('y'); ax.legend()
plt.show()

import matplotlib.pyplot as plt
import numpy as np

# 准备 2D 网格数据
x = np.linspace(-3, 3, 100)
y = np.linspace(-3, 3, 100)
X, Y = np.meshgrid(x, y)
Z = np.exp(-(X**2 + Y**2) / 4) * np.cos(2*X) * np.sin(2*Y)

fig, ax = plt.subplots(figsize=(6, 5))

# 填色 + 等高线叠加（最常用的科研画法）
cf = ax.contourf(X, Y, Z, levels=20, cmap='RdBu_r')
ax.contour(X, Y, Z, levels=10, colors='black', linewidths=0.5, alpha=0.5)

fig.colorbar(cf, ax=ax, label='Z value')
ax.set_xlabel('X'); ax.set_ylabel('Y')
plt.show()

import matplotlib.pyplot as plt
import numpy as np

data = np.random.randn(20, 30)

fig, axes = plt.subplots(1, 2, figsize=(11, 4))

# imshow：默认 y 轴反向（图像坐标系），用 origin='lower' 改回数学坐标系
im1 = axes[0].imshow(data, cmap='viridis', origin='lower', aspect='auto')
axes[0].set_title('imshow')
fig.colorbar(im1, ax=axes[0])

# pcolormesh：自动配合坐标值
x = np.arange(31); y = np.arange(21)
im2 = axes[1].pcolormesh(x, y, data, cmap='viridis', shading='auto')
axes[1].set_title('pcolormesh')
fig.colorbar(im2, ax=axes[1])

plt.show()

import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import pandas as pd
import numpy as np

# 一年的日数据
dates = pd.date_range('2024-01-01', periods=365)
values = np.cumsum(np.random.randn(365))

fig, ax = plt.subplots(figsize=(10, 4))
ax.plot(dates, values, color='#2E86AB', linewidth=1.5)

# 关键 3 行：定位器 + 格式化器 + 旋转标签
ax.xaxis.set_major_locator(mdates.MonthLocator())          # 每月一个刻度
ax.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m')) # 显示成 YYYY-MM
fig.autofmt_xdate()                                          # 自动旋转防重叠

ax.set_xlabel('Date'); ax.set_ylabel('Value')
plt.show()

import matplotlib.pyplot as plt
import numpy as np

x = np.arange(1, 13)
temp = 15 + 10 * np.sin((x - 4) * np.pi / 6)  # 温度 5~25 °C
rain = np.array([50, 60, 80, 100, 150, 200, 250, 220, 150, 100, 70, 50])  # 降水 mm

fig, ax1 = plt.subplots(figsize=(8, 4.5))

# 左 y 轴：温度
l1 = ax1.plot(x, temp, 'o-', color='#E63946', linewidth=2, label='Temperature')
ax1.set_xlabel('Month')
ax1.set_ylabel('Temperature (°C)', color='#E63946')
ax1.tick_params(axis='y', labelcolor='#E63946')

# 右 y 轴：降水
ax2 = ax1.twinx()
l2 = ax2.bar(x, rain, color='#2E86AB', alpha=0.4, label='Rainfall')
ax2.set_ylabel('Rainfall (mm)', color='#2E86AB')
ax2.tick_params(axis='y', labelcolor='#2E86AB')

# 合并图例
ax1.legend(handles=[l1[0], l2], loc='upper left')
plt.show()

import seaborn as sns
import matplotlib.pyplot as plt

# 一句话设主题（影响全局）
sns.set_theme(
    style='whitegrid',          # 'white' / 'darkgrid' / 'ticks' 等
    palette='colorblind',       # 色盲友好色板
    font_scale=1.1              # 全局字号放大 1.1 倍
)

# 加载内置数据集（练手用）
tips = sns.load_dataset('tips')
print(tips.head())

import seaborn as sns
import matplotlib.pyplot as plt

tips = sns.load_dataset('tips')

fig, axes = plt.subplots(1, 2, figsize=(11, 4))

# histplot：直方图 + KDE 叠加
sns.histplot(data=tips, x='total_bill', kde=True,
             color='#2E86AB', ax=axes[0])
axes[0].set_title('Histogram with KDE')

# kdeplot：按类别分组的密度估计
sns.kdeplot(data=tips, x='total_bill', hue='sex',
            fill=True, alpha=0.4, ax=axes[1])
axes[1].set_title('KDE grouped by sex')

plt.tight_layout()
plt.show()

import seaborn as sns

iris = sns.load_dataset('iris')

# jointplot：两变量联合分布（散点 + 边缘直方图）
sns.jointplot(data=iris, x='sepal_length', y='petal_length',
              hue='species', kind='scatter', height=5)

# pairplot：所有数值列两两关系（对角线是分布）
sns.pairplot(iris, hue='species', diag_kind='kde',
             palette='colorblind', height=2.0)

import seaborn as sns
import matplotlib.pyplot as plt

tips = sns.load_dataset('tips')

fig, axes = plt.subplots(1, 2, figsize=(11, 4))

# boxplot
sns.boxplot(data=tips, x='day', y='total_bill', hue='sex',
            palette='colorblind', ax=axes[0])
axes[0].set_title('Boxplot')

# violinplot：split=True 让两组共享中线（节省空间）
sns.violinplot(data=tips, x='day', y='total_bill', hue='sex',
               split=True, inner='quartile',
               palette='colorblind', ax=axes[1])
axes[1].set_title('Violinplot (split)')

plt.tight_layout()
plt.show()

import seaborn as sns
import matplotlib.pyplot as plt

tips = sns.load_dataset('tips')

fig, axes = plt.subplots(1, 2, figsize=(11, 4))

# regplot：单图回归
sns.regplot(data=tips, x='total_bill', y='tip',
            scatter_kws={'alpha': 0.5, 's': 30},
            line_kws={'color': '#E63946'},
            ax=axes[0])
axes[0].set_title('Linear regression')

# 二次拟合（order=2）
sns.regplot(data=tips, x='total_bill', y='tip', order=2,
            scatter_kws={'alpha': 0.5},
            line_kws={'color': '#06A77D'},
            ax=axes[1])
axes[1].set_title('Quadratic fit')

plt.tight_layout()
plt.show()

import seaborn as sns
import matplotlib.pyplot as plt

iris = sns.load_dataset('iris').select_dtypes('number')
corr = iris.corr()

# 1. 基础热力图：相关矩阵
fig, ax = plt.subplots(figsize=(6, 5))
sns.heatmap(corr, annot=True, fmt='.2f',
            cmap='RdBu_r', center=0,
            vmin=-1, vmax=1, square=True,
            cbar_kws={'shrink': 0.8}, ax=ax)
ax.set_title('Correlation Matrix')
plt.show()

# 2. 聚类热图：自动行列聚类
sns.clustermap(iris, cmap='viridis',
               standard_scale=1,    # 按列标准化
               figsize=(8, 6))

# 安装：pip install scienceplots
import scienceplots
import matplotlib.pyplot as plt
import numpy as np

# 切换到 science 样式（Nature 风格 + 黑白可读）
plt.style.use(['science', 'nature', 'no-latex'])
# 可选样式叠加：'ieee', 'grid', 'high-vis', 'bright', 'dark_background'

x = np.linspace(0, 10, 100)

fig, ax = plt.subplots(figsize=(3.5, 2.5))  # Nature 单栏宽度
ax.plot(x, np.sin(x), label=r'$\sin(x)$')
ax.plot(x, np.cos(x), label=r'$\cos(x)$')
ax.set_xlabel('x'); ax.set_ylabel('y'); ax.legend()
plt.show()

# 安装：pip install missingno
import missingno as msno
import pandas as pd
import numpy as np

# 构造一个有缺失的数据集
rng = np.random.default_rng(42)
df = pd.DataFrame({
    'A': rng.normal(0, 1, 100),
    'B': rng.normal(0, 1, 100),
    'C': rng.normal(0, 1, 100),
    'D': rng.normal(0, 1, 100),
})
# 人为制造缺失
df.loc[rng.choice(100, 20), 'A'] = np.nan
df.loc[rng.choice(100, 30), 'B'] = np.nan
df.loc[rng.choice(100, 10), 'C'] = np.nan

# 三种视角
msno.matrix(df)      # 矩阵图：每列每行哪里缺
msno.bar(df)         # 条形图：每列缺多少
msno.heatmap(df)     # 热力图：列之间的缺失相关性

# 安装：conda install -c conda-forge cartopy
import cartopy.crs as ccrs
import cartopy.feature as cfeature
import matplotlib.pyplot as plt

fig = plt.figure(figsize=(10, 5))

# 关键：subplot 时指定 projection
ax = fig.add_subplot(1, 1, 1, projection=ccrs.Robinson())

# 加地图要素
ax.add_feature(cfeature.COASTLINE, linewidth=0.5)
ax.add_feature(cfeature.BORDERS,   linewidth=0.3, linestyle=':')
ax.add_feature(cfeature.OCEAN,     facecolor='#A8D8EA')
ax.add_feature(cfeature.LAND,      facecolor='#F4F1DE')

ax.set_global()  # 显示全球
ax.gridlines(draw_labels=True, alpha=0.3)
plt.show()

import cartopy.crs as ccrs
import cartopy.feature as cfeature
import matplotlib.pyplot as plt
import numpy as np

# 模拟全球温度场
lon = np.linspace(-180, 180, 73)
lat = np.linspace(-90, 90, 37)
LON, LAT = np.meshgrid(lon, lat)
T = 30 * np.cos(np.deg2rad(LAT)) - 10  # 简化的温度模型

fig = plt.figure(figsize=(11, 5))
ax = fig.add_subplot(1, 1, 1, projection=ccrs.Robinson())

# 等值线填色（注意 transform）
cf = ax.contourf(LON, LAT, T, levels=15, cmap='RdBu_r',
                  transform=ccrs.PlateCarree())

# 散点：几个城市
cities = {'Beijing': (116, 40), 'NY': (-74, 41), 'Sydney': (151, -34)}
for name, (x, y) in cities.items():
    ax.plot(x, y, 'o', color='black', markersize=6,
            transform=ccrs.PlateCarree())
    ax.text(x+3, y+3, name, transform=ccrs.PlateCarree(), fontsize=9)

ax.add_feature(cfeature.COASTLINE, linewidth=0.5)
fig.colorbar(cf, ax=ax, orientation='horizontal',
             label='Temperature (°C)', shrink=0.6, pad=0.05)
plt.show()

# 安装：conda install -c conda-forge geopandas
import geopandas as gpd
import matplotlib.pyplot as plt

# 读 shapefile（这里用 Geopandas 自带的世界国家数据）
world = gpd.read_file(gpd.datasets.get_path('naturalearth_lowres'))
print(world.head())
print(world.columns)  # geometry 列存几何形状

fig, ax = plt.subplots(figsize=(11, 6))

# 按 GDP per capita 着色（choropleth 地图）
world['gdp_per_capita'] = world['gdp_md_est'] / world['pop_est']
world.plot(column='gdp_per_capita', ax=ax,
           cmap='viridis', legend=True,
           legend_kwds={'label': 'GDP per capita',
                        'shrink': 0.6},
           missing_kwds={'color': 'lightgray'})  # 缺失填灰
ax.set_xlabel('Longitude'); ax.set_ylabel('Latitude')
plt.show()

import cartopy.crs as ccrs
import cartopy.feature as cfeature
import matplotlib.pyplot as plt
from matplotlib.patches import FancyArrowPatch

fig = plt.figure(figsize=(8, 8))
ax = fig.add_subplot(1, 1, 1, projection=ccrs.PlateCarree())

# 显示中国范围
ax.set_extent([70, 140, 15, 55], crs=ccrs.PlateCarree())

ax.add_feature(cfeature.COASTLINE, linewidth=0.5)
ax.add_feature(cfeature.BORDERS,   linewidth=0.5)

# 1. 经纬度网格（自带标签）
gl = ax.gridlines(draw_labels=True, alpha=0.3,
                  linewidth=0.5, linestyle='--')
gl.top_labels = False; gl.right_labels = False

# 2. 指北针（简化版）
ax.annotate('N', xy=(0.95, 0.95), xytext=(0.95, 0.88),
            xycoords='axes fraction', ha='center', fontsize=14,
            fontweight='bold',
            arrowprops=dict(arrowstyle='-|>', color='black'))

# 3. 比例尺（手画一条横线 + 文字）
ax.plot([75, 85], [18, 18], 'k-', linewidth=2,
        transform=ccrs.PlateCarree())
ax.text(80, 19, '~1000 km', ha='center', fontsize=9,
        transform=ccrs.PlateCarree())

plt.show()

import plotly.express as px

# 用内置数据集
df = px.data.iris()

# 散点图（一行）
fig = px.scatter(df, x='sepal_width', y='sepal_length',
                  color='species', size='petal_length',
                  hover_data=['petal_width'],   # 悬停显示额外列
                  title='Iris Dataset')
fig.show()

# 箱线图按类别分组
fig = px.box(df, x='species', y='sepal_length', color='species')
fig.show()

# 直方图带分组
fig = px.histogram(df, x='sepal_length', color='species',
                    barmode='overlay', opacity=0.7)
fig.show()

import plotly.graph_objects as go
import numpy as np

x = np.linspace(0, 10, 100)

fig = go.Figure()

# 加迹线
fig.add_trace(go.Scatter(x=x, y=np.sin(x), mode='lines',
                          name='sin', line=dict(color='#2E86AB', width=2)))
fig.add_trace(go.Scatter(x=x, y=np.cos(x), mode='lines+markers',
                          name='cos', line=dict(color='#E63946', width=2, dash='dash'),
                          marker=dict(size=4)))

# 布局
fig.update_layout(
    title='Sin and Cos',
    xaxis_title='x',
    yaxis_title='y',
    template='plotly_white',     # 主题：'plotly' / 'plotly_white' / 'plotly_dark'
    hovermode='x unified',       # 悬停时横向对齐显示所有迹线
    width=800, height=450
)

fig.show()

import plotly.graph_objects as go
from plotly.subplots import make_subplots
import numpy as np

x = np.linspace(0, 10, 100)

# 2×2 子图布局
fig = make_subplots(rows=2, cols=2,
                     subplot_titles=('Sin', 'Cos', 'Tan', 'Random'),
                     vertical_spacing=0.15)

fig.add_trace(go.Scatter(x=x, y=np.sin(x), name='sin'), row=1, col=1)
fig.add_trace(go.Scatter(x=x, y=np.cos(x), name='cos'), row=1, col=2)
fig.add_trace(go.Scatter(x=x, y=np.tan(x), name='tan'), row=2, col=1)
fig.add_trace(go.Scatter(x=x, y=np.random.randn(100).cumsum(),
                          name='walk'), row=2, col=2)

fig.update_layout(height=600, width=900,
                   title_text='Subplots Demo',
                   template='plotly_white')

# 导出
fig.write_html('dashboard.html')               # 完整 HTML（含 Plotly.js，体积大）
fig.write_html('dashboard.html', include_plotlyjs='cdn')  # 用 CDN（体积小）

fig.show()

# 安装：conda install networkx
import networkx as nx
import matplotlib.pyplot as plt

# 创建图
G = nx.Graph()
G.add_edges_from([
    ('A', 'B'), ('A', 'C'), ('B', 'C'), ('B', 'D'),
    ('C', 'D'), ('D', 'E'), ('E', 'F'), ('F', 'A')
])

# 节点位置（布局算法决定美观度）
pos = nx.spring_layout(G, seed=42)   # 弹簧布局，最常用

fig, ax = plt.subplots(figsize=(7, 5))

# 分层画：边 → 节点 → 标签
nx.draw_networkx_edges(G, pos, alpha=0.5, width=1.5)
nx.draw_networkx_nodes(G, pos, node_size=800,
                        node_color='#2E86AB', alpha=0.9)
nx.draw_networkx_labels(G, pos, font_color='white',
                         font_weight='bold')

ax.set_axis_off()
plt.show()

import plotly.graph_objects as go

fig = go.Figure(data=[go.Sankey(
    node = dict(
        pad=15, thickness=20,
        line=dict(color='black', width=0.5),
        label=['Energy', 'Coal', 'Gas', 'Solar',
               'Electricity', 'Loss'],
        color='#2E86AB'
    ),
    link = dict(
        # 第 0 个节点 → 第 1/2/3 个节点
        source = [0, 0, 0, 1, 2, 3],
        target = [1, 2, 3, 4, 4, 4],
        value  = [40, 30, 30, 30, 25, 28],   # 流量大小
    )
)])

fig.update_layout(title='Energy Flow', font_size=12)
fig.show()

from sklearn.datasets import load_digits
from sklearn.decomposition import PCA
from sklearn.manifold import TSNE
import matplotlib.pyplot as plt
import numpy as np

# 加载手写数字（64 维）
digits = load_digits()
X, y = digits.data, digits.target

# 三种降维
X_pca  = PCA(n_components=2).fit_transform(X)
X_tsne = TSNE(n_components=2, random_state=42).fit_transform(X)

# 可视化
fig, axes = plt.subplots(1, 2, figsize=(13, 5))

for ax, X_emb, title in zip(axes, [X_pca, X_tsne], ['PCA', 't-SNE']):
    sc = ax.scatter(X_emb[:, 0], X_emb[:, 1], c=y,
                    cmap='tab10', s=10, alpha=0.7)
    ax.set_title(title)
    ax.set_xlabel('Dim 1'); ax.set_ylabel('Dim 2')
fig.colorbar(sc, ax=axes, label='Digit', ticks=range(10))
plt.show()

# UMAP（pip install umap-learn 后）
# import umap
# X_umap = umap.UMAP(random_state=42).fit_transform(X)

import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(42)
n = 2000

# 模拟差异表达数据
log2fc = rng.normal(0, 1.5, n)
pval = 10 ** -np.abs(rng.normal(0, 1.5, n))

# 阈值
fc_thresh, p_thresh = 1, 0.05

# 分组
neg_log_p = -np.log10(pval)
up    = (log2fc >  fc_thresh) & (pval < p_thresh)
down  = (log2fc < -fc_thresh) & (pval < p_thresh)
notsig = ~(up | down)

fig, ax = plt.subplots(figsize=(7, 6))
ax.scatter(log2fc[notsig], neg_log_p[notsig], c='lightgray', s=8, alpha=0.5, label='NS')
ax.scatter(log2fc[up],     neg_log_p[up],     c='#E63946', s=12, alpha=0.7, label='Up')
ax.scatter(log2fc[down],   neg_log_p[down],   c='#2E86AB', s=12, alpha=0.7, label='Down')

# 阈值参考线
ax.axhline(-np.log10(p_thresh), color='gray', linestyle='--', linewidth=0.8)
ax.axvline( fc_thresh, color='gray', linestyle='--', linewidth=0.8)
ax.axvline(-fc_thresh, color='gray', linestyle='--', linewidth=0.8)

ax.set_xlabel(r'$\log_2$(Fold Change)')
ax.set_ylabel(r'$-\log_{10}$(p-value)')
ax.legend()
plt.show()

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

rng = np.random.default_rng(42)

# 模拟 GWAS 数据：22 条染色体，每条 1000 SNP
chroms = np.repeat(range(1, 23), 1000)
positions = np.tile(np.arange(1000), 22)
pvals = rng.uniform(1e-8, 1, 22000)

df = pd.DataFrame({'chr': chroms, 'pos': positions, 'pval': pvals})

# 累计位置（让 x 轴连续）
df['cum_pos'] = 0
offset = 0
chr_centers = []   # 每条染色体中心点（放标签用）
for c in range(1, 23):
    mask = df['chr'] == c
    df.loc[mask, 'cum_pos'] = df.loc[mask, 'pos'] + offset
    chr_centers.append(offset + 500)
    offset += 1000

# 画图
fig, ax = plt.subplots(figsize=(12, 4))
colors = ['#2E86AB', '#888888']
for c in range(1, 23):
    sub = df[df['chr'] == c]
    ax.scatter(sub['cum_pos'], -np.log10(sub['pval']),
               c=colors[c % 2], s=4, alpha=0.7)

# 显著性参考线
ax.axhline(-np.log10(5e-8), color='red', linestyle='--', linewidth=1)

ax.set_xticks(chr_centers)
ax.set_xticklabels(range(1, 23), fontsize=8)
ax.set_xlabel('Chromosome'); ax.set_ylabel(r'$-\log_{10}$(p)')
plt.show()

from sklearn.metrics import confusion_matrix
import seaborn as sns
import matplotlib.pyplot as plt
import numpy as np

# 模拟预测结果
rng = np.random.default_rng(42)
y_true = rng.integers(0, 3, 500)
y_pred = y_true.copy()
flip = rng.choice(500, 80, replace=False)
y_pred[flip] = rng.integers(0, 3, 80)

cm = confusion_matrix(y_true, y_pred)

fig, ax = plt.subplots(figsize=(5.5, 4.5))
sns.heatmap(cm, annot=True, fmt='d', cmap='Blues',
            xticklabels=['Class A', 'Class B', 'Class C'],
            yticklabels=['Class A', 'Class B', 'Class C'],
            cbar=False, ax=ax)
ax.set_xlabel('Predicted'); ax.set_ylabel('Actual')
plt.show()

from sklearn.metrics import roc_curve, auc, precision_recall_curve
import matplotlib.pyplot as plt
import numpy as np

rng = np.random.default_rng(42)
y_true = rng.integers(0, 2, 500)
y_score = y_true + rng.normal(0, 0.5, 500)  # 模拟模型分数

# 计算
fpr, tpr, _ = roc_curve(y_true, y_score)
precision, recall, _ = precision_recall_curve(y_true, y_score)
roc_auc = auc(fpr, tpr)

fig, axes = plt.subplots(1, 2, figsize=(11, 4.5))

# ROC
axes[0].plot(fpr, tpr, color='#2E86AB', linewidth=2,
              label=f'AUC = {roc_auc:.3f}')
axes[0].plot([0, 1], [0, 1], 'k--', alpha=0.5, label='Random')
axes[0].set_xlabel('False Positive Rate'); axes[0].set_ylabel('True Positive Rate')
axes[0].set_title('ROC Curve'); axes[0].legend()

# PR
axes[1].plot(recall, precision, color='#E63946', linewidth=2)
axes[1].set_xlabel('Recall'); axes[1].set_ylabel('Precision')
axes[1].set_title('Precision-Recall Curve')

plt.tight_layout()
plt.show()

from sklearn.model_selection import learning_curve
from sklearn.ensemble import RandomForestClassifier
from sklearn.datasets import make_classification
import numpy as np
import matplotlib.pyplot as plt

X, y = make_classification(n_samples=1000, random_state=42)

train_sizes, train_scores, val_scores = learning_curve(
    RandomForestClassifier(n_estimators=50, random_state=42),
    X, y, cv=5,
    train_sizes=np.linspace(0.1, 1.0, 10)
)

# 计算均值与标准差（用于误差带）
train_mean = train_scores.mean(axis=1); train_std = train_scores.std(axis=1)
val_mean   = val_scores.mean(axis=1);   val_std   = val_scores.std(axis=1)

fig, ax = plt.subplots(figsize=(7, 4.5))
ax.plot(train_sizes, train_mean, 'o-', color='#2E86AB', label='Train')
ax.fill_between(train_sizes, train_mean - train_std, train_mean + train_std,
                 alpha=0.2, color='#2E86AB')
ax.plot(train_sizes, val_mean, 's-', color='#E63946', label='Validation')
ax.fill_between(train_sizes, val_mean - val_std, val_mean + val_std,
                 alpha=0.2, color='#E63946')
ax.set_xlabel('Training samples'); ax.set_ylabel('Accuracy')
ax.legend(); plt.show()