Constant axes/spine border size for multiple plots

Question

In Python Matplotlib, I would need give advice how to keep exactly same axes/spine border size for multiple plots. Independent of how many digits the axis contains.

Generated plots:

Code bellow shows example of generating two plots, each with different axis value (1 digit vs. 3/4 digit on axis), axis+spine border size is different. Respective in case lower digit count is border larger, for higher digit count it is sligtly smaller. I would need to keep same size. Thank you for help

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

plt.style.use('classic')
plt.rcParams.update({'mathtext.default': 'regular'})
plt.rcParams['font.family'] = 'Arial'
plt.rcParams["axes.linewidth"] = 0.50


i=0

plots=['plot1','plot2']

xlabel=['x_plot1','x_plot2']
ylabel=['y_plot1','y_plot2']

for i in range(len(plots)):
    fig, ax = plt.subplots( figsize=[6.0, 4.8])
    fig.tight_layout()

    if(i==0):
        ax.plot([0,9], [0,9])
        ax.axis([0, 9, 0, 9])

    if(i==1):
        ax.plot([0,1000], [0,1000])
        ax.axis([0, 1000, 0, 1000])


    ax.set_xlabel(xlabel[i],fontsize='12')
    ax.set_ylabel(ylabel[i],fontsize='12')
    
    ax.grid(linestyle='solid', linewidth=0.6, color='#000000', which='major', zorder=1, clip_on=False)  #New #EAEAEA   old #C3C3C3
    plt.gcf().canvas.draw()
    ax.set_axisbelow(True)
    ax.spines['bottom'].set_linewidth(1.5)
    ax.spines['left'].set_linewidth(1.5)

    plt.yticks(fontsize='12', zorder=0)
    plt.xticks(fontsize='12', zorder=0)

    plt.savefig(plots[i]+'.png', dpi=300, bbox_inches='tight', transparent=False)
    plt.clf()

Tried change figsize value, aspect ratio of graphs, others, but no satisfactory results.

I expect set static value where should be place axes in figure, create enough margin for more digits on axis.

Answer 1

This is expected behavior when passing bbox_inches='tight' to plt.savefig: the plot margins will be adjusted to fit the axes, but since the width of your tick labels vary, it comes to different results. Remove bbox_inches='tight' and specify the layout with subplots_adjust:

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

plt.style.use('classic')
plt.rcParams.update({'mathtext.default': 'regular'})
plt.rcParams['font.family'] = 'Arial'
plt.rcParams["axes.linewidth"] = 0.50


i=0

plots=['plot1','plot2']

xlabel=['x_plot1','x_plot2']
ylabel=['y_plot1','y_plot2']

for i in range(len(plots)):
    fig, ax = plt.subplots( figsize=[6.0, 4.8])
    fig.tight_layout()

    if(i==0):
        ax.plot([0,9], [0,9])
        ax.axis([0, 9, 0, 9])

    if(i==1):
        ax.plot([0,1000], [0,1000])
        ax.axis([0, 1000, 0, 1000])


    ax.set_xlabel(xlabel[i],fontsize='12')
    ax.set_ylabel(ylabel[i],fontsize='12')
    
    ax.grid(linestyle='solid', linewidth=0.6, color='#000000', which='major', zorder=1, clip_on=False)  #New #EAEAEA   old #C3C3C3
    plt.gcf().canvas.draw()
    ax.set_axisbelow(True)
    ax.spines['bottom'].set_linewidth(1.5)
    ax.spines['left'].set_linewidth(1.5)

    plt.yticks(fontsize='12', zorder=0)
    plt.xticks(fontsize='12', zorder=0)
    plt.subplots_adjust(0.2, 0.1, 0.9, 0.9)

    plt.savefig(plots[i]+'.png', dpi=300, transparent=False)
    plt.clf()

Output: