tlancon / cim Goto Github PK

For portability, the saved script should remember landmarks.

Maybe instead, implement saving of projects in #5 and just save a companion project for all saved napari scripts

Opening a project should:

• Reopen the reference and moving images that were saved
• Remember any registrations computed for image pairs
• Remember landmark color selections
• Remember landmark locations for each image

Opening a project should not be allowed unless the user answers "Yes" to a "Do you understand this will clear all current unsaved work" prompt.

Saving will just create the file needed for opening, so can work backwards.

JSON dumping/reading is probably a good way to handle this.

This extra .py file can be a place for the user to dump extra functionality that will be imported into napari when the overlay script is opened.

For example

Scripts generated by CIM now cause the following when executed:

The 'gui_qt()' context manager is deprecated.
If you are running napari from a script, please use 'napari.run()' as follows:


    import napari

    viewer = napari.Viewer()  # no prior setup needed
    # other code using the viewer...
    napari.run()

In IPython or Jupyter, 'napari.run()' is not necessary. napari will automatically
start an interactive event loop for you:

    import napari
    viewer = napari.Viewer()  # that's it!

  warn(

Currently there is a list of landmarks for each moving image, and corresponding landmarks for the reference image saved WITH each moving image.

This is good because changing the moving image also changes which reference landmarks are displayed.

This causes an issue because it means if you want to reuse the same landmarks on the reference image for another image (or another hundred images?????) you have to redo them for each moving image.

Add ability to copy reference landmarks across moving images to get around this

Does matplotlib provide a way to do this easily/quickly?