科研数据可视化

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

# 调用可视化技能
scientific-skills:matplotlib

# 示例：年龄和血压的散点图
# 输入："用matplotlib画一个散点图，
#        展示年龄和血压的关系，
#        添加回归线和置信区间"

# 生成的代码
plt.figure(figsize=(8, 6))
sns.regplot(data=df, x='age', y='bp',
            scatter_kws={'alpha':0.5},
            line_kws={'color':'red'})
plt.xlabel('Age (years)', fontsize=12)
plt.ylabel('Blood Pressure (mmHg)', fontsize=12)
plt.title('Age vs Blood Pressure', fontsize=14)
plt.tight_layout()
plt.savefig('scatter_plot.pdf', dpi=300)

import matplotlib.pyplot as plt
import seaborn as sns

# 调用可视化技能
scientific-skills:seaborn

# 示例：三组患者生存时间比较
# 输入："用seaborn画一个箱线图，
#        比较三组患者的生存时间，
#        添加显著性标注（*p<0.05, **p<0.01）"

# 生成的代码
plt.figure(figsize=(8, 6))
ax = sns.boxplot(data=df, x='group', y='survival_time',
                     order=['Control', 'Treatment A', 'Treatment B'])

# 添加显著性标注
from statannotations import Annotator
pairs = [(('Control', 'Treatment A'), '**'),
         (('Control', 'Treatment B'), '***'),
         (('Treatment A', 'Treatment B'), 'ns')]
annotator = Annotator(ax, pairs, data=df, x='group', y='survival_time')
annotator.configure(test='Mann-Whitney').apply_and_annotate()

plt.ylabel('Survival Time (days)', fontsize=12)
plt.tight_layout()

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

# 调用可视化技能
scientific-skills:seaborn

# 示例：10个变量的相关性热图
# 输入："用seaborn画一个相关性热图，
#        展示10个变量之间的相关系数，
#        用红蓝色阶，标注数值"

# 计算相关系数矩阵
corr_matrix = df.corr()

# 绘制热图
plt.figure(figsize=(10, 8))
sns.heatmap(corr_matrix,
            annot=True,           # 显示数值
            cmap='RdBu_r',       # 红蓝配色
            center=0,             # 中心值为0
            fmt='.2f',            # 保留两位小数
            cbar_kws={'label': 'Correlation'})
plt.title('Correlation Matrix', fontsize=14)
plt.tight_layout()

import plotly.graph_objects as go
from lifelines import KaplanMeierFitter

# 调用可视化技能
scientific-skills:plotly

# 示例：交互式生存曲线
# 输入："用plotly画一个交互式生存曲线，
#        可以悬停显示详细信息，
#        导出为HTML"

# 计算KM曲线
kmf = KaplanMeierFitter()
fig = go.Figure()

for group in df['treatment'].unique():
    group_data = df[df['treatment'] == group]
    kmf.fit(group_data['time'], group_data['event'], label=group)
    fig.add_trace(go.Scatter(
        x=kmf.survival_function_.index,
        y=kmf.survival_function_['KM_estimate'],
        mode='lines',
        name=group,
        hovertemplate='%{x}天: 生存率%.2f%%<extra></extra>'
    ))

fig.update_layout(
    title='Kaplan-Meier Survival Curve',
    xaxis_title='Time (days)',
    yaxis_title='Survival Probability',
    hovermode='x unified'
)

# 导出为HTML
fig.write_html('survival_curve.html')