使用正确的轴刻度标签将百分比轴添加到 Seaborn catplot答案

【问题标题】：Add percentage axis to Seaborn catplot with correct axis tick labels使用正确的轴刻度标签将百分比轴添加到 Seaborn catplot
【发布时间】：2021-01-22 17:52:46
【问题描述】：

我正在绘制一个分类变量的计数，并想添加第二个 y 轴来显示样本总数的百分比。

import matplotlib.pyplot as plt
import seaborn as sns

titanic = sns.load_dataset("titanic")

g = sns.catplot(x="alive", col="embark_town", col_wrap=4,
                data=titanic[titanic.deck.notnull()],
                kind="count", height=4, aspect=.8)

for i, ax in enumerate(g.axes.flat):
    
    # Create second y-axis for the percentages on the right
    ax1 = ax.twinx()
    
    ### Attempt to fix percentages by plotting the bars over
    #g = sns.catplot(x="alive", col="embark_town", col_wrap=4,
    #            data=titanic[titanic.deck.notnull()],
    #            kind="count", height=4, aspect=.8,
    #            ax = ax1)
    
    # Label by the percentages
    ax1.set_ylim(ax.get_ylim())
    ax1.set_yticklabels(np.round(ax.get_yticks()/titanic[titanic.deck.notnull()].shape[0],1))
    ax1.set_ylabel('Percentage')

    # Rotate x-labels
    labels = ax.get_xticklabels() # get x labels
    ax.set_xticklabels(labels, rotation=90)
  
# Ensure good spacing  
g.fig.tight_layout()

好的，现在我的问题是百分比在右侧 y 轴上重复，如下图所示

我试图通过在新轴上绘制计数来纠正此问题，但这会添加另一行子图（请参阅 for 循环中注释掉的代码）。如何获得正确的 y 轴标签以不包含重复值并实际反映总数的百分比？

【问题讨论】：

您的标签显示为“重复”，因为您将比率四舍五入到小数点后一位（只需将1 更改为3 即可查看差异）。也许将它们乘以100，例如像这样：percent = (ax.get_yticks() / titanic[titanic.deck.notnull()].shape[0] * 100).astype('int') 和 ax1.set_yticklabels(percent) `
哈哈哈好收获！

标签： python-3.x matplotlib seaborn

【解决方案1】：

import matplotlib.pyplot as plt
import seaborn as sns

titanic = sns.load_dataset("titanic")

g = sns.catplot(x="alive", col="embark_town", col_wrap=4,
                data=titanic[titanic.deck.notnull()],
                kind="count", height=4, aspect=.8)

# calculate numbre of samples:
total_samples = len(titanic[titanic.deck.notnull()])


for i, ax in enumerate(g.axes.flat):
    
#     bounds of the left y-axis:
    ymin, ymax = ax.get_ylim()    
    
#     # Create second y-axis for the percentages on the right    
    ax1 = ax.twinx()
    
    # scale right axis labels to total samples and mutliply with 100 for percentages
    ax1.set_ylim(100*ymin/total_samples, 100*ymax/total_samples)
    
# Ensure good spacing  
g.fig.tight_layout()

【讨论】：