Matplotlib | Error bars on line charts and scatter plots (errorbar)

Experimental data are subject to basic errors, which are indicated by error bars on the graph.

Python can easily display a graph with error bars with little code!

This article describes in detail how to display error bars on line charts and scatter plots in Matplotlib.

From displaying graphical error bars on a line chart, to applying them to a scatter plot, to detailed Axes.errorbar function arguments.

Error bars on a line chart (Axes.errorbar)

In Matplotlib, the Axes.errorbar function displays an error bar.

Error bars in array (yerr)

The simplest line chart with error bars.

The error bar data is defined in an array.

The following tabs explain the code and flowchart

import matplotlib.pyplot as plt
import numpy as np

# step1 Create data
x = np.arange(5)
y = [4.6, 3.6, 5, 4.2, 2.7]
yerr = [0.6, 0.9, 1.2, 0.5, 0.7]

# step2 Create graph frames
fig, ax = plt.subplots()

# step3 line chart with error bars
ax.errorbar(x, y, yerr)

ax.set_xlabel('X label')
ax.set_ylabel('Y label')
ax.set_title('Simple errorbar')

plt.show()

Error bars in scalar (xrr, yerr)

The simplest line chart with error bars.

The error bar data is defined by a single number only.

# step3 line chart with error bars
ax.errorbar(x, y, xerr=0.2, yerr=0.4)

Error bars on scatter plots and their format (fmt)

fmt (format string) can specify marker, linetype, and color at once as follows

fmt = '[marker][line][color]'

It is also possible to make a scatter plot without lines and with markers only

Three types of graphs were created using fmt

# step3 line chart with error bars
# Circle + green + scatter plot
ax.errorbar(x, y, yerr=0.4, fmt='og' ,label='fmt="og"')
# Rectangle + red + single dotted line
ax.errorbar(x, [i+2 for i in y], yerr=0.4, fmt='sr-.' ,label='fmt="sr-."')
# Star + cyan + dotted line
ax.errorbar(x,  [i+4 for i in y], yerr=0.4, fmt='*c:' ,label='fmt="*c:"')

Color and thickness of error bars

This section describes how to define the color and thickness of error bars

Color of error bars (ecolor)

The color of the error bar is set using the ecolor argument to the Axes.errorbar function

In the following, I used ecolor=’red’ to make it red

If ecolor=None, the color will be the same as the marker

# step3 line chart with error bars
ax.errorbar(x, y, yerr=0.4, fmt='o-', ecolor='red')

Thickness of error bars (elinewidth)

The thickness of the error bar is determined by the linelinewidth argument to the Axes.errorbar function

In the following I used linewidth=’0.5

# step3 line chart with error bars
ax.errorbar(x, y, yerr=0.4, fmt='o-', elinewidth=0.5)

Error bar caps

This section describes how to define the length and thickness of the cap that represents the width of the error bar.

Also, here is an explanation of how to turn the cap into an arrow

Length of cap (capsize)

The length of the caps on the error bar is determined by the capsize argument of the Axes.errorbar function

In the following, capsize=0 to eliminate the cap portion

# step3 line chart with error bars
ax.errorbar(x, y, yerr=0.4, fmt='o-', ecolor='red', capsize=0)

Thickness of cap (capthick)

The thickness of the error bar cap is determined by the capthick argument of the Axes.errorbar function

In the following, capthick=5 is used to make the cap thicker

# step3 line chart with error bars
ax.errorbar(x, y, yerr=0.4, fmt='o-', ecolor='red', capthick=5)

Change y-direction caps to arrows (uplims, lolims)

The uplims or lolims argument to the Axes.errorbar function can be used to change the caps on the error bar to arrows

In the following, uplims is the positive side in the y direction and lolims is the negative side in the y direction

# step3 line chart with error bars
ax.errorbar(x, y, yerr=0.4, fmt='o-', uplims=True, label='uplims')
ax.errorbar(x, [i+2 for i in y], yerr=0.4, fmt='o-', lolims=True, label='lolims')

Change x-direction caps to arrows (xuplims, xlolims)

The xuplims or xlolims argument to the Axes.errorbar function can be used to change the caps on the error bar to arrows

In the following, xuplims is the positive side in the x direction and xlolims is the negative side in the x direction

# step3 line chart with error bars
ax.errorbar(x, y, xerr=0.2, fmt='o-', xuplims=True, label='xuplims')
ax.errorbar(x, [i+2 for i in y], xerr=0.2, fmt='o-', xlolims=True, label='xlolims')

Error bar display style

This section describes the arguments related to the display of error bars

Define the z-position of the error bar, frequency of display, and arrow display.

z-position of error bar (barsabove)

The z-position of error bars and markers can be changed with barsabove

The argument barsabove=True in the Axes.errorbar function makes the error bar appear before the marker

# step3 line chart with error bars
ax.errorbar(x, y, yerr=0.4, fmt='o-', ecolor='gray', barsabove=True)

Frequency of error bar display (errorevery)

The frequency of displaying the error bar is determined by the argument of the Axes.errorbar function: errorrevery

In the following, error bars are displayed every 2 data with errorrevery=2

# step3 line chart with error bars
ax.errorbar(x, y, yerr=0.4, fmt='o-', ecolor='red', errorevery=2)

References

Axes.errorbar function

How to write the error bar