Floor the log axis lower bound when an explicit margin is set

[Sanjays2402]

Closes #350

The explicit-margin log branch of Figure._range() computed out_lo = 10.0 ** (transformed_lo - pad) with no floor, so a wide-but-legal domain (1e-300 to 1e10 with margin=0.1) pushed the padded exponent past the smallest representable double and the lower bound underflowed to 0.0, leaving a log axis with a non-positive lower bound. This applies the same np.nextafter(0.0, 1.0) floor the default-margin branch already uses; the lo / 10.0 clamp is deliberately not carried over, since it would override an authored margin.

Regression test added in tests/test_figure.py beside test_log_autorange_uses_positive_zone_stats; it fails on main (assert np.float64(0.0) > 0.0) and passes with the fix. tests/test_figure.py is otherwise baseline-identical (15 pre-existing failures locally, all headless-browser PNG export).

This change was prepared with AI assistance; the regression test was run locally and fails without the fix.

Summary by CodeRabbit

• Bug Fixes
  • Fixed log-scale charts with explicit margins potentially producing a zero or invalid lower bound.
  • Ensured the lower bound remains positive while preserving a valid chart range.

[coderabbitai]

📝 Walkthrough

Walkthrough

The explicit-margin log-axis calculation now prevents the lower bound from underflowing to zero. A regression test covers extreme data magnitudes and verifies a positive, ordered y-range.

Changes

Log-axis margin handling

Layer / File(s)	Summary
Positive lower-bound protection python/xy/_figure.py, tests/test_figure.py	The explicit-margin log-axis lower bound is floored at np.nextafter(0.0, 1.0), with a regression test confirming the lower bound is positive and below the upper bound.

Estimated code review effort: 1 (Trivial) | ~5 minutes

Suggested reviewers: alek99, farhanaliraza

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Linked Issues check	✅ Passed	The fix applies the positive floor in the explicit-margin log branch and adds a regression test for the underflow case.
Out of Scope Changes check	✅ Passed	The changes stay focused on the reported log-axis underflow fix and its regression test, with no unrelated scope visible.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately summarizes the main fix: flooring the log-axis lower bound when an explicit margin is configured.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@tests/test_figure.py`:
- Around line 1692-1698: The test
test_log_explicit_margin_keeps_lower_bound_positive should verify the exact
positive minimum-float floor, not just that lo is positive. Add an assertion
that lo equals np.nextafter(0.0, 1.0), while retaining the existing hi > lo
check.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: bdbb4482-c00e-46ff-a515-8ae15c8afb0f

📥 Commits

Reviewing files that changed from the base of the PR and between bd1d36e and 0bad3dd.

📒 Files selected for processing (2)

• python/xy/_figure.py
• tests/test_figure.py

[coderabbitai]

🎯 Functional Correctness | 🟡 Minor | ⚡ Quick win

Assert the intended floor, not only positivity.

This test would also pass with the prohibited lo / 10.0 clamp. Assert that lo == np.nextafter(0.0, 1.0) (or otherwise verify lo < 1e-300) so the regression test proves the explicit-margin path uses the positive minimum-float floor.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@tests/test_figure.py` around lines 1692 - 1698, The test
test_log_explicit_margin_keeps_lower_bound_positive should verify the exact
positive minimum-float floor, not just that lo is positive. Add an assertion
that lo equals np.nextafter(0.0, 1.0), while retaining the existing hi > lo
check.

[cubic-dev-ai]

1 issue found across 2 files

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="tests/test_figure.py">

<violation number="1" location="tests/test_figure.py:1697">
P3: This assertion only checks `lo > 0.0`, which would also pass under the rejected `lo / 10.0` clamp approach. To actually prove the explicit-margin path uses the `np.nextafter(0.0, 1.0)` floor rather than some other positive clamp, assert the exact expected value (e.g. `lo == np.nextafter(0.0, 1.0)`) or a tight upper bound like `lo < 1e-300`.</violation>
</file>

<sub>Reply with feedback, questions, or to request a fix.

Fix all with cubic | Re-trigger cubic</sub>

[cubic-dev-ai]

P3: This assertion only checks lo > 0.0, which would also pass under the rejected lo / 10.0 clamp approach. To actually prove the explicit-margin path uses the np.nextafter(0.0, 1.0) floor rather than some other positive clamp, assert the exact expected value (e.g. lo == np.nextafter(0.0, 1.0)) or a tight upper bound like lo < 1e-300.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At tests/test_figure.py, line 1697:

<comment>This assertion only checks `lo > 0.0`, which would also pass under the rejected `lo / 10.0` clamp approach. To actually prove the explicit-margin path uses the `np.nextafter(0.0, 1.0)` floor rather than some other positive clamp, assert the exact expected value (e.g. `lo == np.nextafter(0.0, 1.0)`) or a tight upper bound like `lo < 1e-300`.</comment>

<file context>
@@ -1689,6 +1689,15 @@ def test_log_autorange_uses_positive_zone_stats():
+    fig = Figure().scatter(np.array([0.0, 1.0]), np.array([1e-300, 1e10]))
+    fig.set_axis("y", type_="log", margin=0.1)
+    lo, hi = fig.y_range()
+    assert lo > 0.0
+    assert hi > lo
+
</file context>

Suggested change

	assert lo > 0.0
	assert lo == np.nextafter(0.0, 1.0)