Skip to content

Tutorial: export figures

Use this workflow for a publication figure, presentation, supplement, or a consistent image set for downstream processing.

1. Select the view

Choose the participant and trial. Keep only layers that answer the question:

  • Fixations + saccades + text for a conventional scanpath;
  • Heatmap + text for spatial concentration;
  • Compare for two trials of the same text;
  • Animate for a talk or supplement.

Set the monitor size first, under Data Management → Edit dataset → Recording setup; it defines the figure's coordinate system.

2. Make one clean figure

Check the whole canvas for clipped marks, unreadable text, and an unnecessary legend. When several figures must share identical grid marks, set a manual interval under Figure & canvas → Axes & grid → Grid (untick Auto).

Open Export → Current figure and choose:

Need Format
editable vector SVG or PDF
image for slides/web PNG
interactive inspection HTML
replay HTML, GIF, or MP4

PNG and SVG are saved by your browser from the figure on screen, and HTML needs nothing either (it loads Plotly from the internet when opened, or tick Self-contained HTML for a larger file that opens offline); PDF, GIF and MP4 need Chrome, Chromium or Edge. For print, give the PNG a Width (mm or in) and a DPI: 180 mm at 600 dpi is 4,252 px wide, and Share → Code writes the same size. Without a width, the plot's own camera button saves the same PNG.

3. Export a batch when needed

In Export → Export bundle:

  1. choose this trial, the active filtered pool, or the whole dataset;
  2. select figure and table formats;
  3. preview the filename pattern;
  4. start the export and inspect one file before using the batch.

For post-production, enable separable layers so text, boxes, fixations, saccades, heatmap, and stimulus image can be stacked in a vector editor.

4. Keep provenance

Keep plot_config.json with the batch (or a Share → File settings file for a single figure). The bundle's README.md records the package version and which trials it was built from, and index.csv lists every file with its participant, trial and screen, and any that failed. Record the dataset version and the trial filters you used in the caption or analysis log — the export does not store them.

Done: the exported files share one visual configuration and can be recreated. For scripted runs, see Automation.