Check journal figures with AI
Find print-size readability issues and trace some of them to the plotting code. Keep publisher requirements separate from visual suggestions.
Check journal figures with AI before sending a revised manuscript to your co-authors. A small legend that is readable in your plotting window may disappear when the journal reduces the figure to one column. With elsevier-figure-style you can specify the final width and ask the agent to inspect the actual export at that size.
The workflow can also revise or generate quantitative plots from source data. For a mechanism diagram or graphical abstract, it reviews the supplied artwork. Findings distinguish publisher guidance from editorial experience and general visual judgement.
What it does
- Identify a faint line or unclear legend at the intended print width.
- Inspect bitmap resolution against the type of artwork you plan to submit.
- Check that panel labels sit consistently across a multi-panel figure.
- Adjust an existing matplotlib plot and keep colours that encode a meaningful threshold.
- Ask for a visual review of a graphical abstract before commissioning changes.
Install and use
- For the Python tools, prepare Python 3.10 or later and the dependencies identified by the environment checker. The R plotting route uses ggplot2 and jsonlite.
- Provide an export, the target journal and the expected figure width. Include plotting code if available.
- Ask, “Check this two-panel result at single-column size. Separate export problems from changes I should make in the plotting script.”
- Review the issue locations and reopen the revised figure at the same size. Confirm the current requirements for your specific journal.
FAQ
Source
Made by Peng Ji and contributors. Original skill: elsevier-figure-style, published on GitHub under the MIT licence.
View on GitHub