Avoid symmetrizing diagonal in hermitian power

[jishnub]

This would make these operations work with immutable matrices as well. The Hermitian wrapper ensures that the result has a real diagonal.

[dkarrasch]

Agreed. Anyone reading from the parent directly must be careful anyway.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.75%. Comparing base (95703b5) to head (3749bbb).
Report is 1 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1321      +/-   ##
==========================================
+ Coverage   93.73%   93.75%   +0.01%     
==========================================
  Files          34       34              
  Lines       15743    15734       -9     
==========================================
- Hits        14757    14751       -6     
+ Misses        986      983       -3     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[araujoms]

Shouldn't this be done with the other matrix functions as well for consistency?

[jishnub]

Are there ones where such a change is necessary? In general, yes, we want the wrappers to work with immutable parents as well.

[araujoms]

sincos also does it. I was sure that sqrt and log also did but it seems that somebody already removed it.