V2 pint unit management - #1684

Codecov Report

❌ Patch coverage is 84.96732% with 23 lines in your changes missing coverage. Please review.
✅ Project coverage is 12.31%. Comparing base (fbde549) to head (fa1a4ee).
⚠️ Report is 1 commits behind head on main.

@@            Coverage Diff             @@
##             main    #1684      +/-   ##
==========================================
+ Coverage   12.07%   12.31%   +0.24%     
==========================================
  Files         116      117       +1     
  Lines       19159    19221      +62     
  Branches     4483     4485       +2     
==========================================
+ Hits         2313     2368      +55     
- Misses      16763    16770       +7     
  Partials       83       83              

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

This is great. The one thing IMO we need to decide on is where exactly to attach the units in the parsers. My initial thought was right at the parse location (utils.float(line.split()[0]) * ureg.... etc.) but there is something to be said for in the return value like return {scfenergies.__name__: [constructed_data] * ureg.hartree}. That way is cleaner but further from where the data was parsed, esp. if it's in non-standard units.

Yes this is true. I have introduced one change, but i think there may be another.

• I made one change which is instantiating the list with units as the array is created.
• What might be missing is attaching units to the specific parsed value that results from this entry into the specific parser.

# initialize empty array with units attached
>>> a = np.array([])*ureg.hartree
>>> a
<Quantity([], 'hartree')>

# adding unit tagged option of same unit is a clean process.
>>> b = 4.5*ureg.hartree
>>> b
<Quantity(4.5, 'hartree')>
>>> np.append(a,b)
<Quantity([4.5], 'hartree')>

# if for any reason different units come up during parsing, this can still be cleanly kept. (can't think of a use case, but just what's possible)
>>> c = 3.5*ureg.eV
>>> c
<Quantity(3.5, 'electron_volt')>
>>> np.append(a,c)
<Quantity([0.12862263], 'hartree')>

From in person conversations, I am just attaching native units from the given program and we will determine when to convert to cclib units at a different point, but it's not clear where that will happen yet.