You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, score expressions are evaluated by converting the scores on a layer into a flat matrix. For example, a layer with two annotated techniques with the scores 42, 51 would be converted into [42, 51]. The score-expression evaluation is performed on this matrix. This leads to confusion since the requirement for elementwise operators (e.g .* instead of *) is not intuitive (#79).
Rather than documenting this behavior as proposed by #80, instead we should change the way things are calculated to be elementwise by default. This will lead to a more intuitive experience and also allow for expressions like conditionals (e.g a > 40? a : b) which are currently not possible.
The text was updated successfully, but these errors were encountered:
Currently, score expressions are evaluated by converting the scores on a layer into a flat matrix. For example, a layer with two annotated techniques with the scores 42, 51 would be converted into [42, 51]. The score-expression evaluation is performed on this matrix. This leads to confusion since the requirement for elementwise operators (e.g .* instead of *) is not intuitive (#79).
Rather than documenting this behavior as proposed by #80, instead we should change the way things are calculated to be elementwise by default. This will lead to a more intuitive experience and also allow for expressions like conditionals (e.g a > 40? a : b) which are currently not possible.
The text was updated successfully, but these errors were encountered: