Integrate time coordinate with mobility functions

[elbeejay]

This PR does the time integration for the mobility functions (relates to #75).

We adopt the same syntax as the section cuts, and simply offer both a basevalues and basevalues_idx input option to set the base maps by either their time, or time index. The same is offered for window and window_idx where the time lag over which mobility functions are computed can be set by either a dimensional value or value representing number of indices.

The dimension of the returned object from the mobility functions should be flexible, so it will correspond to the first dimension of the 3-D xarray (if provided). This would allow someone to apply these methods to any arbitrary model output or input data. Overall a lot of input type handing and conversion goes on to accept inputs that are lists of arrays, or 3-D arrays (and xarray counterparts).

The channel_presence function has also been updated to keep track of dimensions and accept a wider set of input types.

Tests and doc-strings have been updated to reflect changes, and an example has been added to the main page of the mobility API documentation as the computation required to generate a bunch of masks for each individual function docstring seemed excessive. For the singular example we compute land masks and channel masks for 10 of the golfcube time slices - I don't think this is too excessive for the docs build, but we can scale that back a bit if needed.

of note: the masking functionality preserves the time dimension when using __init__ pathways but not via the static methods, this has been noted in #75 and will be the subject of a future PR. Mobility calculation using the _idx inputs will still work if the time dimension is messed up.

edit: additional documentation addresses #29

---

old notes

This PR will eventually fully integrate the time coordinate with the mobility functions...

to do

• switch mobility functions to initialize xarrays instead of ndarrays - inputs already sanitized to become dimensional xarrays if they are not already
• update docstrings to support new set of inputs (basevalues, basevalues_idx, window, window_idx)
• update and add more tests to cover the new cases when input values are not indices but actual time values
• create some sort of example in documentation using the golf-example and the mobility functions (maybe in index page so the set of masks are only generated once)
• flexible coordinate (aka dont assume it is 'time')

[codecov]

Codecov Report

Merging #92 (34fb8d0) into develop (2e9c28a) will increase coverage by 0.02%.
The diff coverage is 92.37%.

@@             Coverage Diff             @@
##           develop      #92      +/-   ##
===========================================
+ Coverage    88.23%   88.25%   +0.02%     
===========================================
  Files            9        9              
  Lines         3060     3109      +49     
===========================================
+ Hits          2700     2744      +44     
- Misses         360      365       +5     

Impacted Files	Coverage Δ
deltametrics/mobility.py	94.50% <92.37%> (-1.74%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 2e9c28a...34fb8d0. Read the comment docs.