Naming and rewrite the very old ones

[MaceKuailv]

No description provided.

[MaceKuailv]

issue #36 is taken care of.
There is also some improvement on naming (issue #33).
itertools.combinations are used instead of home made combination (issue #35 )
This version should pass all tests, but zenodo is done...

[codecov]

Codecov Report

❗ No coverage uploaded for pull request base (main@9205d40). Click here to learn what that means.
Patch has no changes to coverable lines.

Additional details and impacted files

@@           Coverage Diff           @@
##             main      #38   +/-   ##
=======================================
  Coverage        ?   93.15%           
=======================================
  Files           ?       11           
  Lines           ?     2587           
  Branches        ?      705           
=======================================
  Hits            ?     2410           
  Misses          ?       75           
  Partials        ?      102           

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[malmans2]

Nice!
pylint is another good tool to enforce code standards.
I suggest to run it on your code.
It's a very strict tool, so you shouldn't necessarily try to address all the issues reported by pylint.

[MaceKuailv]

All global and eval eliminated, which means issue #16 resolved

[MaceKuailv]

Used an attributeable_dictionary object to refactor many grid operations. Also added a place to include documentation for what I mean by rel coordinate.
Consider issue #35 resolved.

[malmans2]

A couple of thoughts:

1. Are you sure you need the AttributableDict? At first glance it seems a bit over-complicated. Maybe attrs can do the job, or you can just simplify the code? attrs is a very robust package, so if it's helpful you can have it as a mandatory dependency (it's possible that one of your dependencies already uses it).
2. Looks like you use "dunder" methods in various places. You shoud be very careful, and only use those when you really need it. An example: to get the length of a list, do len([1]) rather than [1].__len__()
3. Instantiated classes should be snake_case and not CamelCase. e.g.: ds = xr.Dataset()

[MaceKuailv]

For 1, I do need the update method of dictionary to make it more readable.
For 2, I will refactor the __dict__ today with vars(). However, I don't know what to do with the setattr methods.
For 3, variables like HRel are actually subclasses.

[MaceKuailv]

I am happy with the style of the convention. consider issue #34 resolved.

[MaceKuailv]

I am relatively happy with the current naming convention and readability in general. For phase 1, I consider issue #33 and finally issue #38 resolved. Also, it has become too big a branch.

@malmans2 Comments?

[malmans2]

Looks good!

[malmans2]

A couple of minor things I noticed:

1. pandas is a mandatory dependency of xarray, so you don't need the try/except + get rid of the "hidden" import as _pd
2. do not raise bare Exceptions, they are very hard to catch. The most common is ValueError, but take a look at all exceptions available if you think that there's something better for you use. Alternatively you can create your own exception, but I would only do that if you are going to catch it somewhere. E.g.:

class SeaDuckError(Exception):
    ...
    pass

[MaceKuailv]

I think I did get rid of the Exceptions, except for one of them in a compiled function. except IndexError simply does not work. There should be a workaround. The wifi is down today. I will merge this first and work on the issues you mentioned from a new branch.