Matplotlib生成等尺寸多PDF图形的问题求助

The issue you're facing is that Matplotlib automatically adjusts layout elements like margins based on tick labels, axis titles, and text length—even when using the same figsize. This changes the actual size of your plotting area, making alignment in LaTeX impossible. To fix this, we need to lock the size and position of the plotting axes across all figures, instead of relying on automatic layout adjustments.

Here's how to resolve it step by step:

Key Fixes to Implement #

1. Disable automatic layout: Turn off figure.autolayout so we can manually control margins without Matplotlib overriding them.
2. Set consistent fixed margins: Use plt.subplots_adjust() to define identical left/right/bottom/top margins for every figure. This ensures the core plotting area (where your lines are drawn) stays the exact same size.
3. Standardize label formatting: Keep tick rotation, font size, and formatting consistent across all plots to avoid unexpected layout shifts.

Modified Working Code #

Here's your example code updated to enforce perfectly aligned plotting areas:

from matplotlib.lines import Line2D
import matplotlib.ticker as mtick
import matplotlib.pyplot as plt
from matplotlib import rc
# Activate LaTeX text rendering
rc('text', usetex=True)
from matplotlib import rcParams

# Disable automatic layout (we'll handle margins manually)
rcParams.update({'figure.autolayout': False})

CHUNK_COUNT = 50
common_figsize=(6,5)
plot_counter = 5
x_step = int(CHUNK_COUNT / 10)
new_xticks = list(range(x_step, CHUNK_COUNT + x_step, x_step)) + [1]

# Define FIXED margins for ALL figures (adjust these values if needed)
# Values represent fractions of the total figure width/height
UNIFORM_MARGINS = {
    'left': 0.15,
    'right': 0.95,
    'bottom': 0.15,
    'top': 0.95
}

##### Plot Figure 1
fig, ax = plt.subplots(figsize=common_figsize)
plt.ylabel("Summary of a simple YY axis")
plt.yticks(rotation=45, fontsize=9)  # Consistent font size
ax.yaxis.set_major_formatter(mtick.PercentFormatter(is_latex=False))

for i in range(0, plot_counter):
    xvals = range(1, CHUNK_COUNT + 1)
    yvals = [j + i for j in xvals]
    plt.plot(xvals, yvals)

plt.xticks(new_xticks)
ax.set_xlim(left=1, right=CHUNK_COUNT)
# Apply fixed margins to lock plotting area size
plt.subplots_adjust(**UNIFORM_MARGINS)
plt.savefig("Figure_1.pdf", transparent=True)
plt.close(fig)

##### Plot Figure 2
fig, ax = plt.subplots(figsize=common_figsize)
plt.ylabel("Summary of another YY axis")
plt.yticks(rotation=45, fontsize=9)  # Match font size from Figure 1
ax.yaxis.set_major_formatter(mtick.PercentFormatter(is_latex=False))

for i in range(0, plot_counter):
    xvals = range(1, CHUNK_COUNT + 1)
    yvals = [(j + i)/100 for j in xvals]
    plt.plot(xvals, yvals)

plt.xticks(new_xticks)
ax.set_xlim(left=1, right=CHUNK_COUNT)
# Apply the EXACT same margins as Figure 1
plt.subplots_adjust(**UNIFORM_MARGINS)
plt.savefig("Figure_2.pdf", transparent=True)
plt.close(fig)

Why This Works #

• By setting fixed margins with subplots_adjust(), we reserve a consistent amount of space for labels, ticks, and titles. The remaining area (your actual plotting axes) will have identical dimensions across all figures.
• Disabling figure.autolayout prevents Matplotlib from overriding our manual layout settings based on variable text lengths.

Additional Tips #

• If long LaTeX labels cause overlap, tweak the left margin in UNIFORM_MARGINS (use the same value for all figures) or reduce the font size consistently.
• Avoid using bbox_inches='tight' in savefig()—it will override your fixed margins and reintroduce layout inconsistencies.
• For even more precision, you can use matplotlib.gridspec.GridSpec to define exact axes dimensions, but subplots_adjust() is simpler for most use cases.

内容的提问来源于stack exchange，提问作者Blitzkoder