Was testing out the new isp-workbook-parser (v2.9.0) and discovered an (existing) fuzzy match issue in ISPyPSA.. (specifically noticed matching new entrant technologies against , technology specific location factors - though perhaps comes up elsewhere). The function used was _best_fuzzy_match (in src/ispypsa/templater/helpers.py:139)
The specific case was "Battery Storage (4hrs storage)" - which is equidistant from two different matches:
- "Battery storage (4hrs storage)" - one character different (S instead of s)
- "Battery Storage (8hrs storage)" - one character different (4 instead of 8)
So a tie, which randomly matches.. Makes me wonder about fuzzy matching more generally, but maybe could add in a check for case-insensitive match before the fuzzy match?
Also another candidate typo to tell AEMO about.
Was testing out the new
isp-workbook-parser(v2.9.0) and discovered an (existing) fuzzy match issue inISPyPSA.. (specifically noticed matching new entrant technologies against , technology specific location factors - though perhaps comes up elsewhere). The function used was_best_fuzzy_match(insrc/ispypsa/templater/helpers.py:139)The specific case was "Battery Storage (4hrs storage)" - which is equidistant from two different matches:
So a tie, which randomly matches.. Makes me wonder about fuzzy matching more generally, but maybe could add in a check for case-insensitive match before the fuzzy match?
Also another candidate typo to tell AEMO about.