如何将 y 轴标签添加到 df.boxplot

Question

I generate a boxplot out of a dataframe with:

bplot = df.boxplot(
    column=column_name,
    by=list1,
    fontsize=8,
    layout=(2,1),
    rot=90,
    figsize=(25,20)
)

Where df is my dataframe, column_name is my wanted column and list1 is the list to use for groupby?

I want to label my y-axis with a string. Tried this:

plt.ylabel('ylabel')

but does not work.

Answer 1

DataFrame.boxplot() function returns Axes object. So, instead of using bplot.ylabel('ylabel') which is the function for pyplot object, try to use:

bplot.set_ylabel('ylabel')

If you still got errors, make sure to write the line right after you call the boxplot() function.

As an alternative, you could also create pyplot subplots first which generates the axes object that you need. And then you can specify the axes object on the boxplot function arguments:

from matplotlib import pyplot as plt
fig, ax = plt.subplot()
ax = df.boxplot(
    column=column_name,
    by=list1,
    fontsize=8,
    layout=(2,1),
    rot=90,
    figsize=(25,20),
    ax=ax
)

Answer 2

You need first to define the pd.DataFrame.boxplot() return_type parameter as 'axes' . It will return a pandas.Series with single value containing the matplotlib.axes._subplots.AxesSubplot object. You can then manipulate this object like any matplotlib axes object:

bbplot = df.boxplot(column=column_name,
                    by=list1,
                    fontsize=8,
                    layout=(2,1),
                    rot=90,
                    figsize=(25,20),
                    return_type='axes')

bbplot[0].set_ylabel('yaxis')

Answer 3

尝试这个：

bplot.set_ylabel('ylabel')