Fix sparse grid tests

[pfebrer]

This PR handles the previous problems that the sparse grid had when reducing orbital products because of a mismatch in nsc between the orbital values and weights (e.g. DM). This resulted in trying to access arrays outside of their memory. This PR:

• Introduces a check to see if the shape of the weights array provided is consistent with the nsc of the sparse grid.
• When computing the density from the DM, adds a check to set nsc of the matrix to the sparse grid one.
• Added the possibility of truncating the orbital values in the sparse grid with the nsc of the geometry. This is done by an extra truncate_with_nsc argument in Geometry._orbital_values.
• Change the nsc in sparse grid tests so that it works.
• Reset pre-compute as the default method for density computation, since it is 100x faster (note: I think the direct method could be optimized to be a bit faster).

[codecov]

Codecov Report

Attention: Patch coverage is 69.23077% with 4 lines in your changes missing coverage. Please review.

Project coverage is 87.06%. Comparing base (0cc03dc) to head (55dead8).
Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/sisl/_core/geometry.py	50.00%	2 Missing ⚠️
src/sisl/_sparse_grid.py	50.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #843      +/-   ##
==========================================
+ Coverage   86.68%   87.06%   +0.37%     
==========================================
  Files         402      402              
  Lines       51887    51894       +7     
==========================================
+ Hits        44980    45180     +200     
+ Misses       6907     6714     -193     

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