matplotlib subplot -颜色是交替的

Question

我这里有一个可以打印各种子图的代码。然而，第二个子图的颜色总是交替的。我如何解决这个问题，使所有的颜色都一致？

As you can see, the second subplot has its colors opposite of the first and third. This is consistent for every column

hrlist = [hrdata2015, hrdata2016, hrdata2017]
titles = ["2015", "2016", "2017"]
columns = ["Sex","Education Level","Salary Plan","Grade",
       "Contract Type","Citizenship", "Division"]


for h in columns:

    plt.figure(figsize=(40,40))
    j = 0


    for i in range(len(hrlist)):
        j +=1
        plt.subplot(2,2,j)

        ax1 = sns.countplot(data=hrlist[i],x= h,hue="HR Status", order = hrlist[i][h].value_counts().index)
        ax1.set_title(titles[i])
        ax1.legend(loc = "upper right", prop={'size': 12})
        if(h=="Education Level" or h=="Grade"):
            plt.xticks(fontsize = 9)
        elif (h == "Division"):
            plt.xticks(rotation = 60, fontsize = 8)
        else:
            plt.xticks(fontsize = 12)
        for p in ax1.patches:
            height = p.get_height()
            ax1.text(p.get_x()+p.get_width()/2,
                 height + 1,
                 '{:1.0f}'.format(height,0),
                 ha="center",rotation=0) 


    plt.tight_layout()
    plt.subplots_adjust(top=0.948,
                    bottom=0.115,
                    left=0.052,
                    right=0.986,
                    hspace=0.533,
                    wspace=0.128)
    plt.show()

Answer 1

我的猜测是，hrdata2016的第一行恰好是“非活动的”，而hrdata2015和hrdata2017的第一行都是“活动的”。由于您没有定义色调(颜色)顺序，因此使用了DataFrame中的顺序。按hue_order参数定义色调(颜色)顺序，如下所示：

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

df1 = pd.DataFrame([['F', 'Active'],
                    ['M', 'Inactive'],
                    ['F', 'Inactive'],
                    ['M', 'Active'],
                    ['F', 'Inactive'],
                    ['M', 'Inactive'],
                    ['F', 'Active']], columns=['Sex', 'HR Status'])
df2 = df1.drop(0)
df3 = df2.drop(1)
hrlist = [df1, df2, df3]

h = 'Sex'
for i in range(len(hrlist)):
    plt.subplot(2,2,i+1)
#     ax1 = sns.countplot(data=hrlist[i], x=h, hue="HR Status", order=hrlist[i][h].value_counts().index)
    ax1 = sns.countplot(data=hrlist[i], x=h, hue="HR Status",
                        order=hrlist[i][h].value_counts().index,
                        hue_order=hrlist[i]["HR Status"].value_counts().index)
plt.show()

​

​