Skip to content

Notebook small cleanups from the 17c and #249 reviews #252

Closed

Description

@guijacquemet

These are low-severity items reported by the reviews of PR #250 (17c) and PR #249. None of them changes a result.

  1. PIXEL_CAP docstring: it says the cap covers "all its planes and colours" of one picture. For an RGB header with 4096² planes, a picture reads 3 × 16.7 M pixels. Memory stays bounded, because each plane is within 64 MiB. Fix the docstring, or count the colours in the cap.
  2. Duplicated note in display_panel: the non-finite note is repeated on a normal panel with NaNs ("1 non-finite pixels shown in magenta; 1 non-finite pixels shown in magenta"), and it does not use the plural helper. The new tiles make this visible.
  3. Step 1 read failures: the sentences in the preview cells include the absolute file path. Show the path relative to the experiment folder instead.
  4. Check and save Enlarge view: replicates_detail is not cleared when the replicate boxes are rebuilt, so an old Enlarge view can stay open.
  5. Reader warnings are not shown: for example W_UNUSED_CHANNEL when labels were declared and every folder later became RGB. The notebook never shows the reader's warnings (§4.3 says they are valid, not errors). Consider listing them on Check and save.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions