Improve rand tests

[vprusso]

Started off doing some improvements for the testing in rand/ and hit a few additional areas. Namely:

• Adding parameterization to a subset of tests
• Reformatting lines based on the new longer linting standard
• Fixing both inner_product and outer_product functions (these most likely can be removed as they are just numpy wrappers)
• Adding more comprehensive docstrings to vectors_from_gram_matrix and gram_matrix_to_vectors functions.
• Minor simplifications and cleanup.
• Increasing coverage of tensor product function.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (08ea5ee) 98.0% compared to head (e00c182) 98.1%.

Additional details and impacted files

@@          Coverage Diff           @@
##           master    #265   +/-   ##
======================================
  Coverage    98.0%   98.1%           
======================================
  Files         291     291           
  Lines        6914    6908    -6     
  Branches      786     774   -12     
======================================
- Hits         6781    6777    -4     
  Misses         90      90           
+ Partials       43      41    -2     

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

[vprusso]

Sorry for all the noise on this one. I'll be leaving this one alone for now so it stays stable. Should be good to be reviewed, @purva-thakre !

[vprusso]

Just a ping for this PR and #271 to see if we want to get these in before some of the other PRs coming down the pipe (might make the merging easier if we chuck these in first) fyi @purva-thakre

[purva-thakre]

LGTM except for a few minor changes.

I think we do want to use pytest.raises to test if an error is raised correctly. One thing I noticed after working on some of the issues in #217 is, if numpy is used to check things like this, the covered line might show up as uncovered to pytest.

[vprusso]

Thanks again for the review, @purva-thakre !