Using the Modelica Individuals and the QUDT individuals, auto-generate as much of the mapping CSV as possible. 290 matches have been found, and are often one-to-many (one qudt unit --> many modelica unit classes). The mapping is being done by the script build_auto_mappings.py, and its results are pasted (currently by hand) into mapping-to-qudt-auto.csv. This file is designed to be complemented by mapping-to-qudt-manual.csv.
Two approaches are currently used:
Try to match symbols exactly
Try replacing "." (mo) with " " (qudt) to account for multiplication symbol differences
Try replacing "^" (qudt) with "" (mo) to account for differences in representing exponent
Much smarter matching techniques can and should be used. A complicating factor is the variety of notation styles used in QUDT. For example, multiplication could be represented by a space, a dash, or two symbols placed next to each other.
Current Shortcomings:
QUDT prefixes (e.g.: milli, kilo) are being compared and matched as if they were units. They should be excluded from this exercise.
Using the Modelica Individuals and the QUDT individuals, auto-generate as much of the mapping CSV as possible. 290 matches have been found, and are often one-to-many (one qudt unit --> many modelica unit classes). The mapping is being done by the script build_auto_mappings.py, and its results are pasted (currently by hand) into mapping-to-qudt-auto.csv. This file is designed to be complemented by mapping-to-qudt-manual.csv.
Two approaches are currently used:
Much smarter matching techniques can and should be used. A complicating factor is the variety of notation styles used in QUDT. For example, multiplication could be represented by a space, a dash, or two symbols placed next to each other.
Current Shortcomings: