Minor issues in PV data from database

[dfulu]

Pull Request

Description

Two small changes

• Stop the ID datatype being changed from integer to float during encoding
• Make sure for each (ID, timestamp) pair we only have one row for the power output. The previous version of dropping duplicates of (ID, timestamp, PV output) was causing an error in the table pivot when accessing the development database. There seem to be multiple rows for each (ID, timestamp) in that database.

Checklist:

• [x] My code follows OCF's coding style guidelines
• [x] I have performed a self-review of my own code
• [x] I have made corresponding changes to the documentation
• [ ] I have added tests that prove my fix is effective or that my feature works
• [x] I have checked my code and corrected any misspellings

[codecov[bot]]

Codecov Report

Merging #218 (b44a802) into main (ba5b318) will decrease coverage by 0.05%. The diff coverage is 80.00%.

@@            Coverage Diff             @@
##             main     #218      +/-   ##
==========================================
- Coverage   81.44%   81.40%   -0.05%     
==========================================
  Files         128      128              
  Lines        5568     5570       +2     
==========================================
- Hits         4535     4534       -1     
- Misses       1033     1036       +3     

Files Changed	Coverage Δ
ocf_datapipes/load/pv/live.py	96.29% <80.00%> (-0.70%)	:arrow_down:

... and 2 files with indirect coverage changes

:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more