Fix StackVertical._fix_limits to reliably push joined subplot tick labels inside frame edges

[Copilot]

StackVertical._fix_limits is supposed to expand y-axis limits so tick labels never bleed across the shared frame line into an adjacent panel, but it had three bugs that made the adjustment unreliable or entirely inoperative.

Bugs fixed

• Reversed condition for upper-limit check — yticks[-2] < 1.0 - dy should be >. The top-edge adjustment almost never fired, leaving tick labels on the upper edges of middle/bottom subplots free to overflow into the panel above.

• Fragile heuristic tick indexing — yticks[1] / yticks[-2] assumed the auto-locator always places outermost ticks just outside the visible frame. Replaced with an explicit visibility-window search (-0.01 ≤ ta ≤ 1.01 in axes units) to find the true extremal visible ticks.

• Auto-locator feedback loop — after widening ylim, AutoLocator recalculated ticks and could place a new tick at the enlarged boundary, recreating the overflow. A FixedLocator is now installed whenever limits are changed, freezing tick positions.

Algebraic formulas replace approximate transforms

The old code approximated new limits using the pre-adjustment transform, giving slightly wrong results. The new code solves the placement equations directly:

# Bottom-only (top subplot):  place T_bottom at axes position dy
new_lower = (t_bottom_data - dy * ylim[1]) / (1.0 - dy)

# Top-only (bottom subplot):  place T_top at axes position 1 - dy
new_upper = ylim[0] + (t_top_data - ylim[0]) / (1.0 - dy)

# Both edges (middle subplot): symmetric solution
total_range = (t_top_data - t_bottom_data) / (1.0 - 2.0 * dy)
new_lower = t_bottom_data - dy * total_range
new_upper = t_top_data   + dy * total_range

Log-scale axes skipped

Applying the linear-mapping formulas to a log-scale axis produced non-positive limits and a matplotlib UserWarning. Non-linear y-scales are now detected and skipped early in _fix_limits.

Tests

New test module tests/stonerplots/test_stack_vertical.py covers bottom/top tick clearance per-panel, FixedLocator installation, joined=False (no-op), 2- and 5-panel stacks, and the single-tick edge case.

---

🔒 GitHub Advanced Security automatically protects Copilot coding agent pull requests. You can protect all pull requests by enabling Advanced Security for your repositories. Learn more about Advanced Security.

[codacy-production]

Coverage summary from Codacy

See diff coverage on Codacy

Coverage variation	Diff coverage
✅ -0.03% (target: -1.00%)	✅ 90.91%

Coverage variation details

	Coverable lines	Covered lines	Coverage
Common ancestor commit (ff9cef1)	1022	953	93.25%
Head commit (c249335)	1047 (+25)	976 (+23)	93.22% (-0.03%)

Coverage variation is the difference between the coverage for the head and common ancestor commits of the pull request branch: <coverage of head commit> - <coverage of common ancestor commit>

Diff coverage details

	Coverable lines	Covered lines	Diff coverage
Pull request (#74)	33	30	90.91%

Diff coverage is the percentage of lines that are covered by tests out of the coverable lines that the pull request added or modified: <covered lines added or modified>/<coverable lines added or modified> * 100%

See your quality gate settings    Change summary preferences