As a user, I'd like to configure the main settings of the genetical algorithm so I make it work my own way.
The main section of this page will be a bunch of sliders which will indicate the weights given to each evaluator. We'll have all these evaluators (default percentages given in parenthesys):
Common not consecutive weekdays (32,137% [^percentages]): tries to avoid assigning the same doctor two consecutive weekdays.
Personal not available weekdays (32,137% [^percentages]): tries to avoid assigning a doctor a weekday she has configured as not available. For example, avoiding assigning a doctor a duty on Wednesday if she's configured Wednesdays as not available.
Personal not available dates (32,137% [^percentages]): tries to avoid assigning a doctor a date she has configured as not available; for example, avoiding assigning a doctor November 8th if she's configured that specific date as not available.
Avoid consecutive duty dates -doublets[^doublets]- (1,339% [^percentages]): tries to avoid assigning a doctor two consecutive duty dates; for example, avoiding assigning a doctor November 8th and November 9th.
Fulfill personal desired duty dates (0,134% [^percentages]): tries to assign a doctor a date she has configured as desired duty date.
Fulfill personal desired free dates (0,134% [^percentages]): tries to avoid assigning a doctor a date she has configured as desired free date.
Balance [^balancing] duty dates between doctors (0,107% [^percentages]): tries to balance duty dates between doctors; this is computed using typical deviation between the whole number of duty dates assigned to each one of the doctors.
Balance duty weekend days [^weekend_days] between doctors (0,107% [^percentages]): tries to balance duty weekend dates between doctors (counting only weekend days, see note below); this is computed using typical deviation between the number of weekend duty dates assigned to each one of the doctors.
Balance whole free weekends [^whole_free_weekends] between doctors (0,08% [^percentages]): tries to balance whole free weekends between doctors (see note below for weekend days).
Balance isolated weekend days [^weekend_days] between doctors (0,054% [^percentages]): tries to balance isolated weekend days between doctors (see note below for weekend days).
Maximize whole free weekends [^whole_free_weekends] (0,134% [^percentages]): Try to maximize whole free weekends for doctors (see notes below for weekend days and whole free weekends).
[^doublets]: We call doublet to a pair of consecutive duty dates. For example, assigning a doctor November 8th and November 9th as on-duty.
[^balancing]: Balancing is done trying to minimize the difference between the doctor who has more on-duty days and the doctor who has less.
[^weekend_days]: Weekend days are defined in weekdays.js: by default they are fridays to sundays.
[^whole_free_weekends]: A whole free weekend means that the doctor isn't on-duty in any of the days of the weekend, so she can plan some kind of leisure activity or even trips :smile:
[^percentages]: Current evaluator percentages are default percentages that seem to work well in some hospitals. They could of course be changed depending on doctor's preferences.
As a user, I'd like to configure the main settings of the genetical algorithm so I make it work my own way.
The main section of this page will be a bunch of sliders which will indicate the weights given to each evaluator. We'll have all these evaluators (default percentages given in parenthesys):
[^doublets]: We call doublet to a pair of consecutive duty dates. For example, assigning a doctor November 8th and November 9th as on-duty. [^balancing]: Balancing is done trying to minimize the difference between the doctor who has more on-duty days and the doctor who has less. [^weekend_days]: Weekend days are defined in weekdays.js: by default they are fridays to sundays. [^whole_free_weekends]: A whole free weekend means that the doctor isn't on-duty in any of the days of the weekend, so she can plan some kind of leisure activity or even trips :smile: [^percentages]: Current evaluator percentages are default percentages that seem to work well in some hospitals. They could of course be changed depending on doctor's preferences.