This branch does not change any current behavior**. It enables the possibility of "probabilistic selection" rather than only the current option of pass/fail "binary selection". With these changes, a selector which returns passFactor <- (0,1) changes the default weight of subsequent steps multiplicatively by passFactor. This can be useful for constructing a compatible simulation when thresholds in data vary over different epochs.
\ unless some weird selector is returning floating point numbers on (0,1) instead of bool, int, or None.
elaird
commented
11 years ago
This branch does not change any current behavior**. It enables the possibility of "probabilistic selection" rather than only the current option of pass/fail "binary selection". With these changes, a selector which returns passFactor <- (0,1) changes the default weight of subsequent steps multiplicatively by passFactor. This can be useful for constructing a compatible simulation when thresholds in data vary over different epochs.
\ unless some weird selector is returning floating point numbers on (0,1) instead of bool, int, or None.