Control Matplotlib Axis Ticks with Custom Locators and Formatters

Problem: default ticks crowd or mislabel your data

Matplotlib chooses tick positions automatically. On a wide range or a small figure that choice can produce overlapping labels, uneven spacing, or numbers that are not readable for your domain. The useful takeaway is that tick placement and tick text are separate concerns, both controlled through matplotlib.ticker. A Locator decides where ticks go, a Formatter decides how those values are rendered.

Desired outcome

You want predictable tick intervals, a maximum number of ticks, and human readable labels such as currency or percentages, without changing the plotted data.

Prerequisites

Matplotlib installed in the Python environment where you run the script. A working Axes object from plt.subplots or a figure. Familiarity with ax.xaxis and ax.yaxis. No special permissions are required; changes are in-memory until you save or show the figure.

Core concepts

A Locator is an object that returns a sequence of tick positions for a given view limits. A Formatter receives a numeric value and a position index and must return a string.

Entry points are Axis.set_major_locator and Axis.set_major_formatter. Minor ticks use set_minor_locator and set_minor_formatter. Changing a locator or formatter does not alter data, only its presentation.

Focused procedure

Fixed interval spacing with MultipleLocator

Use MultipleLocator when you need ticks every N units. This is useful for consistent visual grids.

import matplotlib.pyplot as plt
from matplotlib.ticker import MultipleLocator

fig, ax = plt.subplots()
ax.plot([0, 20], [0, 20])
ax.xaxis.set_major_locator(MultipleLocator(base=5))

Expected check: the x axis shows major ticks at multiples of 5 within the current limits. If the limits change, the locator recomputes on draw.

Custom label text with FuncFormatter

FuncFormatter wraps a Python callable that maps value to string. The callable signature is value, position.

from matplotlib.ticker import FuncFormatter

def currency_fmt(x, pos):
    return f'{x:,.0f} USD'

ax.xaxis.set_major_formatter(FuncFormatter(currency_fmt))

Expected check: tick labels are transformed by the function and remain strings. Returning a non-string can cause rendering errors in some backends.

Cap tick count and control precision

MaxNLocator limits the number of major ticks and chooses nice locations. Pair it with FormatStrFormatter for fixed decimal precision.

from matplotlib.ticker import MaxNLocator, FormatStrFormatter

ax.yaxis.set_major_locator(MaxNLocator(nbins=6))
ax.yaxis.set_major_formatter(FormatStrFormatter('%.2f'))

Expected check: no more than about six y ticks are drawn and each label shows two decimal places.

Combined example

import matplotlib.pyplot as plt
from matplotlib.ticker import MultipleLocator, FuncFormatter, MaxNLocator, FormatStrFormatter

def currency_fmt(x, pos):
    return f'{x:,.0f} USD'

fig, ax = plt.subplots()
ax.plot([0, 20], [0, 20])

ax.xaxis.set_major_locator(MultipleLocator(base=5))
ax.xaxis.set_major_formatter(FuncFormatter(currency_fmt))
ax.yaxis.set_major_locator(MaxNLocator(nbins=5))
ax.yaxis.set_major_formatter(FormatStrFormatter('%.1f'))

plt.show()

Risk: too many ticks for the figure DPI and size cause overlapping labels. Reduce nbins or increase figure size.

Limitations and cautions

Logarithmic scales require LogLocator and LogFormatter. Applying a linear MultipleLocator or FuncFormatter to a log axis produces incorrect spacing.

Custom formatters must return a string for every input. Edge cases such as NaN or very large values should be handled explicitly.

Changing locators after the view limits are set is fine, but the change only becomes visible after a draw or show.

Verification

• Instantiate a figure and apply MultipleLocator(base=5). Verify ticks appear at regular 5 unit intervals across the visible range.
• Implement a FuncFormatter that appends a unit symbol. Verify the label transformation logic is applied to each tick.
• Set MaxNLocator with nbins and compare the drawn tick count against the requested maximum.

Recovery options

If a custom locator or formatter produces undesirable results, reset to defaults:

ax.xaxis.set_major_locator(plt.AutoLocator())
ax.xaxis.set_major_formatter(plt.ScalarFormatter())

For a full reset you can also call ax.relim and ax.autoscale_view. These operations are safe and do not modify data.