fix(misc): correct docs wording and visualiser axis-count assertions

[Copilot]

Summary

Applies six requested copy/test fixes: three documentation grammar/spacing corrections and three test assertion tightenings in Captum image visualiser tests. The test updates now validate exact subplot layout expectations instead of permissive lower bounds.

• Docs wording fixes

  • docs/contributor/releases.md: add missing space before toctree after `going-stable-1-0`.
  • docs/index.md: change “then ran” → “then run”.
  • docs/using-raitap/installation.md: change “was currently tested” → “is currently tested”.

• Transparency visualiser test expectations

  • src/raitap/transparency/visualisers/tests/test_visualisers.py
    • test_visualise_tensor: len(fig.axes) == 8 (4 samples × 2 panels).
    • test_max_samples_limit: len(fig.axes) == 8 for max_samples=4.
    • test_can_disable_original_image_panel: len(fig.axes) == 2 when include_original_image=False and max_samples=2.

fig = visualiser.visualise(attributions)
assert len(fig.axes) == 8  # 2 axes per sample (original + attribution) for 4 samples

Checklist

• CI — Required checks are green
• Breaking changes — If this PR breaks compatibility (API, configs, file formats, etc.), the title uses ! after the type or scope (e.g. feat!: or feat(transparency)!:). I understand that this change will have a direct, disruptive impact on package users. I ensured that this change is absolutely necessary and cannot be delayed.
• Contributor guide — I’ve read Pull requests and commit messages before requesting review.

Optional

• Issue — A GitHub issues is linked to this PR ("Development" section in the right sidebar).
• Docs — User or contributor docs updated where needed.
• Tests — New or updated tests cover the change.

Original prompt

Please apply the following diffs and create a pull request.
Once the PR is ready, give it a title based on the messages of the fixes being applied.

[{"message":"Missing space between 'going-stable-1-0' and 'toctree'. Should be `going-stable-1-0` toctree.","fixFiles":[{"filePath":"docs/contributor/releases.md","diff":"diff --git a/docs/contributor/releases.md b/docs/contributor/releases.md\n--- a/docs/contributor/releases.md\n+++ b/docs/contributor/releases.md\n@@ -39,7 +39,7 @@\n \n ## 3. Update the contributor guide\n \n-Delete this file and update {doc}`./index` to remove the `going-stable-1-0`toctree entry.\n+Delete this file and update {doc}`./index` to remove the `going-stable-1-0` toctree entry.\n \n ## 4. Cutting 1.0.0\n \n"}]},{"message":"Corrected verb form from 'ran' to 'run'.","fixFiles":[{"filePath":"docs/index.md","diff":"diff --git a/docs/index.md b/docs/index.md\n--- a/docs/index.md\n+++ b/docs/index.md\n@@ -13,7 +13,7 @@\n \n ## Where does it fit in my workflow?\n \n-RAITAP is configured via YAML [Hydra](https://hydra.cc/) configs or CLI flags, and then ran via a CLI command.\n+RAITAP is configured via YAML [Hydra](https://hydra.cc/) configs or CLI flags, and then run via a CLI command.\n \n This means it can be used either as:\n \n"}]},{"message":"The phrase 'was currently tested' is grammatically inconsistent. Use either 'was tested' (past tense) or 'is currently tested' (present tense).","fixFiles":[{"filePath":"docs/using-raitap/installation.md","diff":"diff --git a/docs/using-raitap/installation.md b/docs/using-raitap/installation.md\n--- a/docs/using-raitap/installation.md\n+++ b/docs/using-raitap/installation.md\n@@ -15,7 +15,7 @@\n ```\n \n :::{note}\n-RAITAP was currently tested with Python 3.13.x. Ensure your project matches this requirement, or expect possible issues.\n+RAITAP is currently tested with Python 3.13.x. Ensure your project matches this requirement, or expect possible issues.\n :::\n \n ## 2. Install optional dependencies\n"}]},{"message":"The comment says 'at least one axes per sample', but the assertion checks for >= 4 axes when sample_images has 4 samples. This doesn't account for the case where include_original_image=True (default), which would create 2 axes per sample (original + attribution). The test should verify 8 axes (2 per sample × 4 samples) to properly validate the paired layout, or explicitly set include_original_image=False if testing single-panel output.","fixFiles":[{"filePath":"src/raitap/transparency/visualisers/tests/test_visualisers.py","diff":"diff --git a/src/raitap/transparency/visualisers/tests/test_visualisers.py b/src/raitap/transparency/visualisers/tests/test_visualisers.py\n--- a/src/raitap/transparency/visualisers/tests/test_visualisers.py\n+++ b/src/raitap/transparency/visualisers/tests/test_visualisers.py\n@@ -40,7 +40,7 @@\n \n         fig = visualiser.visualise(attributions)\n         assert fig is not None\n-        assert len(fig.axes) >= 4  # at least one axes per sample\n+        assert len(fig.axes) == 8  # 2 axes per sample (original + attribution) for 4 samples\n \n     @pytest.mark.usefixtures(\"needs_captum\")\n     def test_max_samples_limit(self) -> None:\n"}]},{"message":"Similar to Comment 1, this assertion checks for >= 4 axes when max_samples=4, but doesn't validate the expected paired layout. With default include_original_image=True and max_samples=4, the figure should have 8 axes total. The test should verify the exact count to ensure proper visualization layout.","fixFiles":[{"filePath":"src/raitap/transparency/visualisers/tests/test_visualisers.py","diff":"diff --git a/src/raitap/transparency/visualisers/tests/test_visualisers.py b/src/raitap/transparency/visualisers/tests/test_visualisers.py\n--- a/src/raitap/transparency/visualisers/tests/test_visualisers.py\n+++ b/src/raitap/transparency/visualisers/tests/test_visualisers.py\n@@ -48,7 +48,7 @@\n         large_batch = torch.randn(64, 3, 32, 32)\n \n         fig = visualiser.visualise(large_batch, max_samples=4)\n-        assert len(fig.axes) >= 4\n+        assert len(fig.axes) == 8\n \n     @pytest.mark.usefixtures(\"needs_captum\")\n     def test_overlay_with_inputs(self, sample_images: torch.Tensor) -> None:\n"}]},{"message":"When include_original_image=False and max_samples=2, this should produce exactly 2 axes (one attribution panel per sample), not >= 2. The test should assert the exact count to verify that the original image panels are indeed disabled.","fixFiles":[{"filePath":"src/raitap/transparency/visualisers/tests/test_visualisers.py","diff":"diff --git a/src/raitap/transparency/visualisers/tests/test_visualisers.py b/src/raitap/transparency/visualisers/tests/test_visualisers.py\n--- a/src/raitap/transparency/visualisers/tests/test_visualisers.py\n+++ b/src/raitap/transparency/visualisers/tests/test_visualisers.py\n@@ -81,7 +81,7 @@\n \n         fig = visualiser.visualise(attributions, inputs=sample_images, max_samples=2)\n \n-        a...

[Copilot]

Pull request overview

This PR applies small documentation wording fixes and tightens Captum image visualiser tests to assert exact Matplotlib subplot/axes layout expectations rather than permissive lower bounds.

Changes:

• Fix minor grammar/spacing issues across three documentation pages.
• Tighten Captum image visualiser tests by asserting specific fig.axes counts for selected scenarios.

Reviewed changes

Copilot reviewed 3 out of 4 changed files in this pull request and generated 2 comments.

File	Description
docs/contributor/releases.md	Fixes missing space in contributor release instructions.
docs/index.md	Corrects wording (“ran” → “run”) in docs homepage.
docs/using-raitap/installation.md	Updates Python version testing sentence to consistent present tense.
src/raitap/transparency/visualisers/tests/test_visualisers.py	Tightens Captum image visualiser axes-count assertions.

[Copilot]

The inline comment claims the 8 axes are from “original + attribution”, but this test calls visualise(attributions) without inputs, so the original-image panel cannot be rendered. Either pass inputs=sample_images (and update the expected axes count accordingly), or update the comment/expectation to reflect what actually creates the extra axes here (typically Captum’s per-sample colorbar axes when show_colorbar=True).

Suggested change

	assert len(fig.axes) == 8 # 2 axes per sample (original + attribution) for 4 samples
	assert len(fig.axes) == 8 # 2 axes per sample for 4 samples: attribution plot + colorbar

[Copilot]

CaptumImageVisualiser defaults to show_colorbar=True, and in the non-paired layout Captum typically adds an extra colorbar axis per sample. With max_samples=2 this usually yields 4 axes (2 plots + 2 colorbars), so len(fig.axes) == 2 is likely incorrect/flaky. To assert exactly 2 axes, set show_colorbar=False for this test, or update the expected axes count to include the colorbar axes.