python用斜率绘制两个时间点

Question

我想把几个图绘制成FacetGrid，每个图包括两个时间点。此外，我还想计算并显示直线的坡度：

ID     TimePoint1    TimePoint2
================================
A      500           20000
B      200           1000
C      3000          50000

大概是这样的：

我尝试了这个代码示例，但图中没有显示任何内容：

import seaborn as sns
import pandas as pd
import matplotlib.pyplot as plt
data = pd.read_table("test.csv", sep=",")
g = sns.FacetGrid(data, col="ID", col_wrap=4, height=2, ylim=(0, 10))
g.map(sns.pointplot, "TimePoint1", "TimePoint2", color=".3", ci=None)

我如何通过seaborn实现这一点

Answer 1

这里使用TimePoint1作为x，TimePoint2作为y。要实现您想要的，您首先需要重塑数据。假设data输入数据：

import pandas as pd
import matplotlib.pyplot as plt
data_long = (data.rename(columns=lambda x: int(x[-1]) if x.startswith('TimePoint') else x)
                 .melt(id_vars=['ID'], var_name='TimePoint')
             )
g = sns.FacetGrid(data_long, col="ID", col_wrap=4, height=2,) # ylim=(0, 10))
g.map(sns.pointplot, "TimePoint", "value", color=".3", ci=None)

输出：

NB。我删除了无效的ylim

数据长度：

  ID TimePoint  value
0  A         1    500
1  B         1    200
2  C         1   3000
3  A         2  20000
4  B         2   1000
5  C         2  50000

您还可以方便地使用^{}，它是FacetGrid的包装器：

sns.catplot(data=df_long, x='TimePoint', y='value', col='ID',
            col_wrap=4, height=2, color=".3", ci=None, kind='point'
           )