Search code examples
pythonmatplotlibseaborn

How to prevent pyplot.errorbar from shifting x-axis of seaborn barplot

I want to plot data using Seaborn barplot; I only have the mean and standard deviation. I use pyplot.errorbar to add error bars to my plot, however, it shifts my x axis slightly (see red star below in plot). How do I prevent this from happening?

Plots: enter image description here

Code to reproduce:

import seaborn as sn
import matplotlib.pyplot as plt 

### loading example data ###
health = sns.load_dataset('healthexp')

health_summary = health.groupby(['Country']).Life_Expectancy.agg({'mean','std'}).reset_index()


### barplot without errorbars ###
p = sn.barplot(health_summary, x = 'Country', y = 'mean', errorbar=None)

plt.show()


### barplot with errorbars ###
p = sn.barplot(health_summary, x = 'Country', y = 'mean', errorbar=None)

p.errorbar(x=health_summary['Country'], y=health_summary['mean'], yerr=health_summary['std'], fmt="none", c="k")

plt.show()
Solution

  • You need to save the xlim before calling errorbar and then restore its values.

    You can make it handy with a context manager:

    from contextlib import contextmanager

import matplotlib.pyplot as plt
import seaborn as sns


@contextmanager
def fixed_xlim(ax):
    xlims = ax.get_xlim()
    try:
        yield
    finally:
        ax.set_xlim(xlims)


health = sns.load_dataset("healthexp")
health_summary = (
    health.groupby(["Country"]).Life_Expectancy.agg({"mean", "std"}).reset_index()
)

fig, (ax1, ax2) = plt.subplots(nrows=2)
p = sns.barplot(health_summary, x="Country", y="mean", errorbar=None, ax=ax1)
p = sns.barplot(health_summary, x="Country", y="mean", errorbar=None, ax=ax2)

with fixed_xlim(ax2):
    p.errorbar(
        x=health_summary["Country"],
        y=health_summary["mean"],
        yerr=health_summary["std"],
        fmt="none",
        c="k",
    )

plt.tight_layout()
plt.show()

    result