Attpc-spyral

This release changes the way that Spyral interprets ion chamber data as well as adding the parsing of FRIBDAQ scalers. Below is a list of major features and their impact.

- FRIBDAQ scalers are now extracted in the pointcloud phase. Scalers are written to their own dataframe and are stored in the scalers directory of the workspace.
- The ion chamber analysis received a bunch of fixes. The multiplicity threshold is now applied more intuitively. The ion chamber delay is now accounted for using a new parameter in the config `ic_delay_time_bucket`. This new parameter allows the user to set the ion chamber delay; all peaks before the delay are discarded and the first peak after (or at) the delay is used as the triggering peak. Also fixed a bug in how the ion chamber offset was applied to point clouds.
- Some small improvements to the progress bars
- Some small bugfixes to notebooks
- Update examine_ic.ipynb for changes to ion chamber.
- Fix some documentation mistakes
- Bump a few deps

This release introduces a new configuration parameter, so please update any existing configuration files you are using. This release also bumps the version of a few dependencies, so please re-install the dependencies in requirements.txt.

This is the v0.3.1 release of Spyral. This is a minor update to address two major edge cases that could cause crashing behavior:

- Spline creation can fail for two major reasons: too few points (less than 5) or the data is not monotonic in z. We now catch these events and drop them, with a warning in the log files.
- On some Linux flavors, multiprocessing defaults to using `fork` instead of `spawn` to start processes. `fork` does not play well with the parallel backend used by polars. So now we strictly require the use of `spawn` to start child processes.

After some testing we also refactored and stabilized the solver_interp code. Unless a revelation occurs, this method most likely will not change dramatically moving forward.

example_analysis.ipynb was also refactored to showcase the use of the vector library bundled as part of spyral-utils.

In this release of Spyral, a major change was made to the way that we perform the estimation phase. We now use smoothing splines to generate a smoothed trajectory for estimating parameters such as dEdx and Brho. This results in significantly improved resolution in the particle ID spectrum in particular!

Below is a brief summary:
- Smoothing splines in estimation. Significant resolution gains in PID at cost of some runtime performance.
- Minor bugfixes in estimation which could result in strange vertex position choices
- Bump versions of many deps. In particular, rocket-fft which was not working for python 3.12.

Since dependencies were updated, please use the requirements.txt to update your local environment after updating!

Documentation is also updated for this release

Forgot to remove the center_x, center_y arguments to solve in phase_solve. Now fixed. No new dependencies or configuration parameters were added to this release.

This release fixes bugs found in the solver fit parameters and the electric field correction (application and generation). No new features, configuration parameters, or dependencies were added to the v0.1.0 code, only these fixes.