pythonviz.blog - Tick formatting recipes

Default tick formatter

from matplotlib.ticker import ScalarFormatter
ax.xaxis.set_major_formatter(ScalarFormatter())

Format string as usually

from matplotlib.ticker import StrMethodFormatter
ax.xaxis.set_major_formatter(StrMethodFormatter('{x:.1f}'))

Ticks in a scientific/engineering notation

from matplotlib.ticker import EngFormatter
ax.xaxis.set_major_formatter(EngFormatter(sep='')) # no unit, no distance to the multiplier 
ax.xaxis.set_major_formatter(EngFormatter(unit='Hz'))

src

Percent formatter that does the scaling too

from matplotlib.ticker import PercentFormatter
value_100perc = 2.5 # data value corresponding to the 100%
ax.xaxis.set_major_formatter(PercentFormatter(xmax=value_100perc))

Lambda formatter (func formatter)

Option 1:

from matplotlib import ticker 

@ticker.FuncFormatter
def custom_formatter(x, pos):
    return f'[{x:.2f}]'

ax.xaxis.set_major_formatter(custom_formatter)

Option 2:

from matplotlib.ticker import FuncFormatter 

custom_formatter = lambda x, pos: f'[{x:.2f}]'
ax.xaxis.set_major_formatter(FuncFormatter(custom_formatter))

src

Space as a separator between thousands

import matplotlib.pyplot as plt 

y = [0, 1e4, 2.3e4, 3.12e4]
x = [2000, 2010, 2020, 2030]

Use the FuncFormatter, which requires a function of a form:

def my_func(x,pos): 
    # blablabla 
    return formatted_x_string

Now implement the formatter. We want every 3 digits separated by a space, and our labels to be integers (won’t work for floats). We will use format(), which accepts parameter ,d producing comma-separated notation, and replace the commas with spaces.

The formatter can be applied to each axis separately.

from matplotlib.ticker import FuncFormatter
format_spaces = lambda x, pos: format(int(round(x)), ',d').replace(',',' ')
ax.yaxis.set_major_formatter(FuncFormatter(format_spaces))

f, ax = plt.subplots()
ax.plot(x,y,marker='o')

from matplotlib.ticker import FuncFormatter
format_spaces = lambda x, pos: format(int(round(x)), ',d').replace(',',' ')
ax.yaxis.set_major_formatter(FuncFormatter(format_spaces))