Figure size for publication in Matplotlib - Deep Dive

Overview - Figure size for publication

What is it?

Figure size for publication refers to setting the exact width and height of a plot or graph so it fits well in printed or digital documents. This ensures the visual looks clear, balanced, and professional when included in articles, reports, or presentations. It involves choosing dimensions that match the publication's requirements, like column width or page size. Proper sizing avoids blurry images or awkward layouts.

Why it matters

Without controlling figure size, plots may appear too large, too small, or distorted in publications, making them hard to read or unprofessional. This can confuse readers or reduce the impact of your data story. Setting the right size helps communicate information clearly and respects the space limits of journals or reports. It also saves time by avoiding repeated resizing and reformatting.

Where it fits

Learners should first understand basic plotting with matplotlib and how to create simple graphs. After mastering figure size, they can learn about resolution (DPI), exporting figures, and styling plots for publication quality. This topic fits early in the journey of preparing visuals for sharing results.

Mental Model

Core Idea

Figure size controls the physical dimensions of a plot to ensure it fits perfectly and looks clear in the final publication.

Think of it like...

It's like choosing the right size frame for a photo before hanging it on a wall; too big or too small looks odd, but the perfect size makes it stand out beautifully.

┌───────────────────────────────┐
│          Figure Size           │
│  (width x height in inches)   │
├───────────────┬───────────────┤
│  Too Small    │  Too Large    │
│  (Unreadable) │ (Wastes space)│
├───────────────┴───────────────┤
│       Just Right (Clear & Fit)│
└───────────────────────────────┘

Build-Up - 7 Steps

1

Foundation

Understanding Matplotlib Figures

Concept: Learn what a figure is in matplotlib and how it holds plots.

In matplotlib, a figure is the overall window or page that contains one or more plots (axes). You create a figure using plt.figure(), which acts like a blank canvas. By default, matplotlib chooses a standard size for this canvas.

Result

You get a blank figure window or image with default size ready for plotting.

Understanding that a figure is the container for plots helps you realize why controlling its size affects the entire visual output.

2

Foundation

Setting Figure Size with figsize

Concept: Learn how to specify figure size using the figsize parameter.

When creating a figure, you can pass figsize=(width, height) in inches. For example, plt.figure(figsize=(6,4)) creates a figure 6 inches wide and 4 inches tall. This controls the physical size of the output image or window.

Result

The figure appears with the specified width and height, changing how much space the plot occupies.

Knowing figsize lets you control the plot's physical dimensions, which is the first step to making publication-ready visuals.

3

Intermediate

Matching Publication Dimensions

🤔Before reading on: Do you think specifying figure size in pixels or inches is better for publications? Commit to your answer.

Concept: Learn why inches are used for figure size and how to match publication column widths.

Publications usually specify figure width in inches or centimeters, not pixels. Using inches with figsize matches these requirements. For example, if a journal column is 3.5 inches wide, set figsize=(3.5, height) to fit exactly. Height can be adjusted to keep aspect ratio or fit layout.

Result

Figures fit perfectly in the publication's column or page width without resizing.

Understanding the unit of inches aligns your figures with real-world print sizes, avoiding guesswork and resizing later.

4

Intermediate

Balancing Aspect Ratio and Readability

🤔Before reading on: Does increasing figure height always improve readability? Commit to your answer.

Concept: Learn how to choose height relative to width to keep plots readable and balanced.

Aspect ratio is width divided by height. A very tall or very wide figure can distort data or make labels cramped. Adjust height to keep plots clear. For example, if width is 6 inches, height between 3 and 4 inches often works well. Test different sizes to find the best balance.

Result

Plots look balanced, with readable labels and clear data presentation.

Knowing how aspect ratio affects perception helps you create visuals that communicate data effectively without distortion.

5

Intermediate

Using DPI to Control Resolution

Concept: Learn how dots per inch (DPI) affects figure clarity and file size.

DPI controls how many pixels per inch the figure has. Higher DPI means sharper images but larger files. You can set DPI with plt.figure(figsize=(w,h), dpi=300) for print quality. Combining figsize and DPI controls both size and sharpness.

Result

Figures have clear details suitable for print or screen, matching publication standards.

Understanding DPI alongside figure size ensures your visuals are both the right size and crisp enough for professional use.

6

Advanced

Automating Figure Size for Multiple Outputs

🤔Before reading on: Can you guess how to create one figure size that works for both screen and print? Commit to your answer.

Concept: Learn techniques to programmatically set figure size and DPI for different output targets.

You can write functions that set figsize and DPI based on output type (screen, print, presentation). For example, use lower DPI and larger size for screen, higher DPI and exact size for print. This automation saves time and ensures consistency across outputs.

Result

You get correctly sized and sharp figures for any medium without manual resizing.

Knowing how to automate sizing and resolution adapts your workflow to real-world publishing needs efficiently.

7

Expert

Handling Complex Layouts with Subplots

🤔Before reading on: Do you think setting figure size alone is enough to make complex subplot layouts publication-ready? Commit to your answer.

Concept: Learn how figure size interacts with subplot spacing and label sizes in multi-plot figures.

When using multiple subplots, figure size must be large enough to fit all plots and their labels without overlap. Use plt.subplots_adjust() to control spacing. Also, font sizes and tick labels may need adjustment to maintain readability. Balancing these factors is key for publication quality.

Result

Complex multi-plot figures look clean, readable, and fit publication dimensions perfectly.

Understanding the interplay between figure size, subplot layout, and label sizing prevents clutter and ensures professional visuals.

Under the Hood

Matplotlib creates figures as objects with a size defined in inches. This size, combined with DPI (dots per inch), determines the pixel dimensions of the output image. When saving or displaying, matplotlib calculates pixel width = inches * DPI and pixel height = inches * DPI. This pixel grid is where all drawing happens. Adjusting figsize changes the canvas size, while DPI changes pixel density, affecting clarity.

Why designed this way?

Using inches and DPI separates physical size from resolution, mirroring how printers and screens work. This design allows precise control over printed size and image sharpness independently. Alternatives like pixel-only sizing would tie size to screen resolution, making print sizing unpredictable. The inch-DPI model aligns with real-world publishing standards.

┌───────────────┐
│  Figure (in)  │  ← Physical size in inches
│  width x height│
└──────┬────────┘
       │ multiplied by DPI
       ▼
┌───────────────┐
│  Pixel Size   │  ← width_px = width_in * DPI
│  width_px x height_px│
└──────┬────────┘
       │ used by backend to render image
       ▼
┌───────────────┐
│  Rendered Plot│  ← final image shown or saved
└───────────────┘

Myth Busters - 4 Common Misconceptions

Quick: Does setting figure size alone guarantee a publication-ready plot? Commit yes or no.

Common Belief: If I set the figure size correctly, my plot will always look perfect in publications.

Reality: Figure size is important but not enough; you must also adjust DPI, font sizes, label spacing, and layout for true publication quality.

Why it matters: Ignoring other factors leads to plots that are the right size but have unreadable text or overlapping elements, reducing clarity.

Quick: Is DPI only about image quality and unrelated to figure size? Commit yes or no.

Common Belief: DPI only affects image sharpness and does not influence the figure's physical size.

Reality: DPI combined with figure size determines the pixel dimensions; changing DPI without adjusting size changes the output resolution and file size.

Why it matters: Misunderstanding DPI can cause images to be blurry or files to be unnecessarily large, wasting storage or printing poorly.

Quick: Can you set figure size in pixels directly in matplotlib? Commit yes or no.

Common Belief: You can specify figure size directly in pixels in matplotlib's figsize parameter.

Reality: Matplotlib's figsize uses inches, not pixels; pixels depend on DPI and inches together.

Why it matters: Trying to set pixels directly leads to confusion and incorrect figure dimensions, especially when preparing for print.

Quick: Does increasing figure size always improve plot readability? Commit yes or no.

Common Belief: Making the figure bigger always makes the plot easier to read.

Reality: Too large figures can cause labels to be spaced too far apart or distort aspect ratio, harming readability and aesthetics.

Why it matters: Blindly increasing size wastes space and can confuse readers with awkward layouts.

Expert Zone

1

Some journals require figures in specific sizes and DPI; knowing how to combine figsize and dpi precisely avoids multiple revisions.

2

When exporting figures, backend differences (e.g., PNG vs PDF) affect how size and DPI are interpreted, requiring format-specific adjustments.

3

Using constrained_layout or tight_layout in matplotlib can help automatically adjust subplot spacing but may need manual tweaks for publication standards.

When NOT to use

Figure size control is not enough when interactive or dynamic plots are needed; in those cases, responsive web-based plotting libraries like Plotly or Bokeh are better. Also, for very complex multi-panel figures, professional design tools like Adobe Illustrator may be preferred for final layout.

Production Patterns

Professionals often create reusable functions that accept target publication width and aspect ratio, automatically setting figsize and DPI. They also save figures in vector formats (PDF, SVG) for scalability and use matplotlib's rcParams to standardize fonts and sizes across projects.

Connections

Image Resolution and DPI

Builds-on

Understanding figure size is incomplete without grasping DPI, as both define the final image quality and size in publications.

Typography in Data Visualization

Complementary

Figure size affects how text elements fit; knowing typography principles helps choose sizes that maintain readability and aesthetics.

Graphic Design Layout Principles

Analogous

Setting figure size for publication parallels graphic design tasks of fitting images into layouts, balancing space, and visual hierarchy.

Common Pitfalls

#1 Setting figure size too large without adjusting DPI, resulting in huge files.

Wrong approach:

plt.figure(figsize=(10,8))
plt.savefig('plot.png') # default dpi=100

Correct approach:

plt.figure(figsize=(10,8), dpi=150)
plt.savefig('plot.png')

Root cause:

Not understanding that default DPI combined with large figsize creates large pixel dimensions and file sizes.

#2 Using figsize in pixels instead of inches, causing unexpected figure sizes.

Wrong approach:

plt.figure(figsize=(800,600)) # incorrect, figsize expects inches

Correct approach:

plt.figure(figsize=(8,6)) # correct, inches

Root cause:

Confusing pixel units with inches in matplotlib's figsize parameter.

#3 Ignoring subplot spacing when increasing figure size, leading to overlapping labels.

Wrong approach:

fig, axs = plt.subplots(2,2, figsize=(8,6))
# no spacing adjustments

Correct approach:

fig, axs = plt.subplots(2,2, figsize=(8,6))
fig.tight_layout()

Root cause:

Assuming larger figure size automatically fixes layout without adjusting spacing.

Key Takeaways

Figure size in matplotlib is set in inches using the figsize parameter and controls the physical dimensions of the plot.

DPI works with figure size to determine the pixel resolution and clarity of the output image.

Matching figure size to publication requirements ensures plots fit perfectly and look professional in printed or digital documents.

Adjusting aspect ratio and subplot spacing alongside figure size is essential for readability and aesthetics.

Understanding these concepts together prevents common mistakes like blurry images, oversized files, or cluttered layouts.