Allow plot-ts to plot SRFs as animations

[lispandfound]

This PR extends plot-ts to allow it to visualise an SRF as an animation (in addition to it's ability to plot xyts files). For example,

$ plot-ts srf alpine_wairau_1.json alpine_wairau_1_modified.srf modified.mp4 --simple-map --fps 5

The old behaviour is obtained with the xyts subcommand

$ plot-ts xyts alpine_wairau_1.json alpine_wairau_1_xyts.e3d modified_xyts.mp4 --simple-map --fps 5

In accordance with the last PR, this SRF plot-ts is not explicitly tested (because I can't think of a good way to do this).

[Copilot]

Pull Request Overview

This PR extends plot-ts to support rendering SRF animations alongside the existing xyts visualisations. Key changes include:

• Introducing a new command for SRF animations (animate_srf_slip_times) along with supporting helper functions.
• Modifying the waveform coordinate computation and reorganising imports for clarity.
• Adjusting the frame rendering logic in both xyts and SRF animation pathways.

Comments suppressed due to low confidence (2)

visualisation/plot_ts.py:293

• The parameter 'ny' is described as 'the number of x-points'; it should likely be corrected to 'the number of y-points' for clarity.

ny : int

visualisation/plot_ts.py:822

• [nitpick] The magic number '20' used for calculating slip_index is hardcoded. Consider defining it as a named constant to explain its significance and facilitate future adjustments.

slip_index = frame_index * 20

[Copilot]

Pull Request Overview

This PR extends the plot-ts utility to support SRF animation rendering alongside the existing XYTS plotting, and refactors waveform coordinate computation. Key changes include:

• Refactoring of xyts_waveform_coordinates to waveform_coordinates with explicit grid parameters.
• Addition of a new Typer CLI command "srf" to animate SRF slip times.
• Rearrangement of certain imports and modification of the frame rendering process in the low frequency animation function.

Comments suppressed due to low confidence (1)

visualisation/plot_ts.py:459

• The render_single_frame function in the low frequency animation now returns a list containing the Axes object instead of saving the generated frame to a file, which may break the subsequent ffmpeg processing expecting image file paths. Consider restoring the functionality to save the frame as an image and returning the corresponding filename.

return [ax]

[lispandfound]

Needed to bump scipy's version to at least 1.15.0 because it needs the sparse indexing operations only supported in the newest stable version.

[Copilot]

Pull Request Overview

This PR extends plot-ts with the ability to animate SRFs while preserving the original xyts plotting functionality. Key changes include refactoring the waveform coordinate computation function, introducing a new CLI command "srf" with its own animation routine, and adding helper functions for processing non-zero data points.

[Copilot]

Pull Request Overview

This PR extends plot-ts to support visualizing SRF files as animations alongside the existing xyts file plotting functionality. Key changes include:

• Addition of a new command (animate_srf_slip_times) for rendering SRF animations.
• Refactoring of waveform coordinate computation to decouple from XYTS-specific logic.
• Adjusted import order and consolidation of redundant import statements.

Comments suppressed due to low confidence (1)

visualisation/plot_ts.py:731

• [nitpick] The attribute name 'slipt1_array' is ambiguous and may be a typo; please verify its intended naming for clarity.

slip = srf_file.slipt1_array.tocsc()

[Copilot]

Pull Request Overview

This PR extends the functionality of plot‑ts by adding support for animating SRF slip times, while still retaining the existing XYTS animation capability.

• Renames and refactors coordinate computation for waveform plotting.
• Introduces a new subcommand for SRF animations with dedicated rendering logic.
• Updates and reorders several imports and adjusts parameter types (dpi, fps) for consistency.

[lispandfound]

The tests are failing due to a bug in openquake. It currently has a dependency requirement of numpy > 1.20, which results in numpy > 2 being installed and openquake doesn't work with numpy > 2. This is all due to a backend requirement in workflow, which visualisation depends on for some plotting utilities related to realisations. The best fix is openquake ports to numpy 2 (ongoing). The second best fix is pinning empirical engine to a numpy version less than 2, which is what I am going to do now.

[joelridden]

Looks good, a rerun of the Jenkins test should work now that Empirical engine has updated and as long as that's passes I'm happy.