Plot Refinement & Publication-Ready Figures

ECON 3209 · Week 5, Lecture 3 · Kerala Agricultural University

Department of Development Economics, KAU

Autumn 2026

Learning Outcomes

By the end of this lecture, you will be able to:

  1. Build multi-panel plots with subplots()
  2. Add formatting, annotations, and reference lines effectively
  3. Save figures for reports and assignments using savefig()
  4. Apply simple design rules for publication-ready visuals
  5. Use live notebook-style plotting to refine figures interactively

What Makes a Plot “Interactive” in Class?

  • In this course, plots are interactive because you can edit code and rerun instantly
  • Small changes in colour, scale, or labels produce immediate visual feedback
  • This makes the notebook a laboratory for visual reasoning
  • Interactivity helps students test economic storytelling choices
  • Publication quality starts with iterative refinement

Building Subplots

  • subplots() creates multiple axes in one figure
  • This helps compare related series side by side
  • Shared formatting gives visual consistency
  • Multi-panel plots are common in reports and presentations

Formatting for Readability

  • Formatting choices shape how quickly an audience understands the figure
  • Remove unnecessary visual noise when possible
  • Rotate crowded tick labels for readability
  • Use consistent fonts, colours, and units across related figures

Adding Reference Lines and Annotations

  • Reference lines help compare data against targets or policy thresholds
  • Annotations can call attention to important events
  • Use them sparingly so the plot remains uncluttered
  • Good annotations direct the audience to the economic message

Saving Figures

  • savefig() exports figures for assignments and reports
  • dpi=300 is a common standard for print-quality output
  • bbox_inches="tight" helps avoid clipped labels
  • Choose meaningful file names so outputs stay organised

Publication-Ready Plot Checklist

  • Clear title with economic meaning
  • Axis labels include units
  • Colour palette is readable and consistent
  • No unnecessary 3D effects or distracting backgrounds
  • Font sizes are large enough for projection and print

Publication-ready does not mean “fancy”; it means accurate, readable, and audience-focused.

Combining Several Design Choices

  • This figure combines labels, grid control, colour consistency, and a clean legend
  • The message is comparative change over time
  • Every design choice should support that message
  • Good figures are built through iteration, not by chance

When to Use Multi-Panel vs Single-Panel Plots

Multi-Panel Helps When

  • Variables have different scales
  • You want side-by-side comparison
  • Several related stories belong together

Single-Panel Helps When

  • One message is central
  • Comparison can be shown clearly in one axes
  • The audience needs a quick takeaway

Common Presentation Mistakes

  • Too many colours with no meaning
  • Tiny labels that are unreadable in class
  • Overlapping text and crowded legends
  • Mixing different units without explanation
  • Saving low-resolution images for final reports

A figure should survive three tests: readable on screen, readable in print, and interpretable without the speaker talking over it.

Exercise

Create a two-panel figure:

  1. Left panel: a line plot of deposits over four years.
  2. Right panel: a line plot of loans over the same years.
  3. Add titles and call plt.tight_layout() before plt.show().

Summary

  • ✅ Live plotting becomes interactive when you edit and rerun code quickly
  • ✅ subplots() helps organise related charts in one figure
  • ✅ Formatting, annotations, and reference lines strengthen interpretation
  • ✅ savefig() exports figures for assignments and reports
  • ✅ Publication-ready figures prioritise clarity over decoration
  • ✅ Strong visuals make econometric arguments easier to follow

Next Lecture

Looking Ahead

  • In the next block of the course we will begin statistical modelling and inference
  • The coding and plotting foundations from Weeks 1–5 will now support regression-based analysis
  • Keep practicing with the live interactive cells before we move onward