是否可以调整 Seaborn 热图中的正方形（单元格）的大小？

Question

假设我有一个像这样的热图 plot：

使用这些数据：

import numpy as np
import pandas as pd


arr = np.array([[ 2,  2,  2,  8,  7,  7,  6,  5,  2,  7,  7,  8,  7,  5,  6,  6,  6],
       [ 8,  7,  5,  4,  4,  3,  9,  6,  7,  4,  3,  2,  8,  9,  3,  3,  3],
       [ 1,  3,  2,  2,  2,  3,  5,  3,  3,  2,  3,  3,  4,  1, 10, 10, 10],
       [ 3,  2,  4,  1,  1,  1,  2,  2,  1,  1,  1,  1,  2,  1,  9,  9,  9],
       [ 7,  6,  7,  6,  6,  6,  2,  2,  5,  6,  5,  4,  7,  9,  9,  9,  9],
       [ 6,  7,  8,  4,  3,  4,  4,  8,  7,  3,  4,  5,  6,  3,  4,  4,  4],
       [ 3,  1,  1,  9,  9,  9,  3,  1,  8,  9,  9,  9,  1,  6,  1,  1,  1],
       [ 3,  3,  3,  5,  5,  5,  5,  1,  2,  5,  6,  5, 10,  8,  8,  8,  8],
       [ 1,  1,  1,  2,  3,  2,  7,  3,  1,  3,  2,  2, 10,  8,  7,  7,  7],
       [ 5,  5,  2,  2,  2,  1,  1,  3,  3,  2,  1,  1,  5,  2,  7,  7,  7],
       [ 7,  9, 10,  3,  4,  4,  8,  9,  9,  3,  4,  6,  2,  3,  2,  2,  2],
       [ 5,  6,  7,  3,  3,  3,  3,  1,  4,  4,  3,  4,  9, 10,  2,  2,  2],
       [ 4,  4,  3,  4,  4,  4,  3,  4,  3,  4,  4,  3,  2,  7, 10, 10, 10],
       [ 2,  1,  1,  8,  8,  8,  1,  4,  2,  8,  8,  8,  4,  1,  5,  5,  5],
       [ 9,  9,  8,  8,  8,  8,  5,  6,  8,  8,  8,  5,  1,  5,  2,  2,  2],
       [ 5,  5,  5,  5,  5,  5,  4,  2,  1,  5,  5,  4,  6,  5,  5,  5,  5],
       [ 8,  8,  9, 10, 10, 10,  6,  7,  6, 10, 10, 10,  3,  7,  4,  4,  4],
       [ 9,  8, 10,  5,  7,  7, 10, 10,  9,  6,  5,  6,  5,  6,  3,  3,  3],
       [10,  9,  9,  7,  6,  5, 10, 10,  9,  8,  7,  8,  3, 10,  8,  8,  8],
       [10, 10,  8, 10, 10, 10,  2,  5, 10, 10, 10,  9,  7,  9,  3,  3,  3],
       [ 4,  4,  5,  3,  2,  2,  9,  8,  4,  2,  2,  3,  4,  4,  5,  5,  5],
       [ 4,  4,  4,  7,  5,  6,  4,  4,  4,  5,  6,  7, 10,  2,  8,  8,  8],
       [ 7,  8,  6,  6,  8,  8,  7,  9,  8,  7,  8,  7,  9,  8,  6,  6,  6],
       [ 8,  7,  7,  7,  7,  7,  8,  9,  5,  7,  7,  7,  5,  7,  1,  1,  1],
       [ 1,  2,  3,  1,  1,  1,  9,  7,  7,  1,  1,  1,  9,  3,  4,  4,  4],
       [ 2,  5,  6,  1,  1,  2,  7,  5,  6,  1,  2,  2,  8,  4,  1,  1,  1],
       [10, 10,  9, 10, 10, 10, 10, 10, 10, 10, 10, 10,  3, 10,  7,  7,  7],
       [ 6,  3,  4,  9,  9,  9,  8,  7,  5,  9,  9, 10,  1,  2, 10, 10, 10],
       [ 9, 10, 10,  9,  9,  9,  1,  8, 10,  9,  9,  9,  8,  4,  9,  9,   9]])

columns = ["feature1", "feature2", "feature3", "feature4", "feature5", "feature6", "feature7", "feature8", "feature9", "feature10", "feature11", "feature12", "feature13", "feature14", "feature15", "feature16", "feature17"]

indexes = ['AAPL', 'AMGN', 'AXP', 'BA', 'CAT', 'CRM', 'CSCO', 'CVX', 'DIS', 'GS',
       'HD', 'HON', 'IBM', 'INTC', 'JNJ', 'JPM', 'KO', 'MCD', 'MMM', 'MRK',
       'MSFT', 'NKE', 'PG', 'TRV', 'UNH', 'V', 'VZ', 'WBA', 'WMT']

df = pd.DataFrame(arr, columns=columns, index=indexes)

使用此代码：

import seaborn as sns
import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(10,10), dpi=600)
a = sns.heatmap(df, annot=True, cmap="RdBu_r", square=True, ax=ax)
plt.show()

我想根据它的值调整每个单元格的大小，我的意思是，值为 1 的方形单元格应该小于具有更高值的单元格！
例子：

请注意，此示例与之前的热图 plot 的值没有严格相关。 我只是提供了一个示例来说明我的意思，即根据其值调整每个方形单元格的大小。

Answer 1

这是您可以使用scatterplot或relplot完成的事情：

flights = sns.load_dataset("flights")
g = sns.relplot(
    data=flights,
    x="year", y="month", size="passengers", hue="passengers",
    marker="s", sizes=(40, 400), palette="blend:b,r",
)

Answer 2

（这篇文章详细阐述了@mwaskom 的优秀解决方案，适用于给定的 dataframe。）

对于大多数 seaborn 功能，将 dataframe 设置为“长格式”会有所帮助。

这是一个示例，说明如何将 dataframe 转换为长格式以获取例如sns.relplot或sns.scatterplot使用的格式。 可能，从用于创建 pivot 表的原始 dataframe 开始会更容易。

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

arr = np.array([[2, 2, 2, 8, 7, 7, 6, 5, 2, 7, 7, 8, 7, 5, 6, 6, 6], [8, 7, 5, 4, 4, 3, 9, 6, 7, 4, 3, 2, 8, 9, 3, 3, 3], [1, 3, 2, 2, 2, 3, 5, 3, 3, 2, 3, 3, 4, 1, 10, 10, 10], [3, 2, 4, 1, 1, 1, 2, 2, 1, 1, 1, 1, 2, 1, 9, 9, 9], [7, 6, 7, 6, 6, 6, 2, 2, 5, 6, 5, 4, 7, 9, 9, 9, 9], [6, 7, 8, 4, 3, 4, 4, 8, 7, 3, 4, 5, 6, 3, 4, 4, 4], [3, 1, 1, 9, 9, 9, 3, 1, 8, 9, 9, 9, 1, 6, 1, 1, 1], [3, 3, 3, 5, 5, 5, 5, 1, 2, 5, 6, 5, 10, 8, 8, 8, 8], [1, 1, 1, 2, 3, 2, 7, 3, 1, 3, 2, 2, 10, 8, 7, 7, 7], [5, 5, 2, 2, 2, 1, 1, 3, 3, 2, 1, 1, 5, 2, 7, 7, 7], [7, 9, 10, 3, 4, 4, 8, 9, 9, 3, 4, 6, 2, 3, 2, 2, 2], [5, 6, 7, 3, 3, 3, 3, 1, 4, 4, 3, 4, 9, 10, 2, 2, 2], [4, 4, 3, 4, 4, 4, 3, 4, 3, 4, 4, 3, 2, 7, 10, 10, 10], [2, 1, 1, 8, 8, 8, 1, 4, 2, 8, 8, 8, 4, 1, 5, 5, 5], [9, 9, 8, 8, 8, 8, 5, 6, 8, 8, 8, 5, 1, 5, 2, 2, 2], [5, 5, 5, 5, 5, 5, 4, 2, 1, 5, 5, 4, 6, 5, 5, 5, 5], [8, 8, 9, 10, 10, 10, 6, 7, 6, 10, 10, 10, 3, 7, 4, 4, 4], [9, 8, 10, 5, 7, 7, 10, 10, 9, 6, 5, 6, 5, 6, 3, 3, 3], [10, 9, 9, 7, 6, 5, 10, 10, 9, 8, 7, 8, 3, 10, 8, 8, 8], [10, 10, 8, 10, 10, 10, 2, 5, 10, 10, 10, 9, 7, 9, 3, 3, 3], [4, 4, 5, 3, 2, 2, 9, 8, 4, 2, 2, 3, 4, 4, 5, 5, 5], [4, 4, 4, 7, 5, 6, 4, 4, 4, 5, 6, 7, 10, 2, 8, 8, 8], [7, 8, 6, 6, 8, 8, 7, 9, 8, 7, 8, 7, 9, 8, 6, 6, 6], [8, 7, 7, 7, 7, 7, 8, 9, 5, 7, 7, 7, 5, 7, 1, 1, 1], [1, 2, 3, 1, 1, 1, 9, 7, 7, 1, 1, 1, 9, 3, 4, 4, 4], [2, 5, 6, 1, 1, 2, 7, 5, 6, 1, 2, 2, 8, 4, 1, 1, 1], [10, 10, 9, 10, 10, 10, 10, 10, 10, 10, 10, 10, 3, 10, 7, 7, 7], [6, 3, 4, 9, 9, 9, 8, 7, 5, 9, 9, 10, 1, 2, 10, 10, 10], [9, 10, 10, 9, 9, 9, 1, 8, 10, 9, 9, 9, 8, 4, 9, 9, 9]])
columns = [f"feature{i}" for i in range(1, 18)]
indexes = ['AAPL', 'AMGN', 'AXP', 'BA', 'CAT', 'CRM', 'CSCO', 'CVX', 'DIS', 'GS', 'HD', 'HON', 'IBM', 'INTC', 'JNJ', 'JPM', 'KO', 'MCD', 'MMM', 'MRK', 'MSFT', 'NKE', 'PG', 'TRV', 'UNH', 'V', 'VZ', 'WBA', 'WMT']
df = pd.DataFrame(arr, columns=columns, index=indexes)
df.index.name = 'Ticker'

df_long = df.reset_index().melt(id_vars='Ticker', var_name='Feature', value_name='Value')
sns.set_style('darkgrid')
g = sns.relplot(data=df_long, x="Feature", y="Ticker", size="Value", hue="Value",
                marker="s", sizes=(20, 200), palette="blend:limegreen,orange", height=8, aspect=1.1)
g.ax.tick_params(axis='x', labelrotation=45)
g.ax.set_facecolor('aliceblue')
g.ax.grid(color='red', lw=1)

g.fig.subplots_adjust(left=0.1, bottom=0.15)
plt.show()