[thatoneguy006]

Fixes #38.

ggforestplot(data, point_size = "N") now maps a numeric source column through ggplot2's size aesthetic instead of passing the column name as a literal marker size. Constant numeric sizes remain supported, and mapped sizes can be customized with scale_size_continuous() or scale_size_area().

Preserve original source-column values and types when canonical normalization converts them to labels, so numeric n columns can also drive marker sizes. Resolve original source names even when they collide with internal columns, and validate sizes only for estimate and summary rows. Missing columns and invalid size values now produce errors before rendering.

Add API documentation, regression tests for sorting, facets, grouped subgroups, row filtering, fitted-model output, source-column collisions, and invalid inputs, plus a reviewed vdiffr snapshot of the split-table workflow from the issue.

Validation:

• R CMD build and R CMD check --no-manual: Status: OK, including vignette builds and checks.
• Complete suite: 851 passed, 0 failures, 0 warnings, 0 skips, including the visual regression snapshots.
• Visually inspected the split table with column-mapped marker sizes and its size legend.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!