Survey Planning Tile Prioritization Rules

Rules for how tiles are prioritized are contained in a YAML file, with a baseline survey example in py/desisurvey/data/rules.yaml. Each tile group is separately scheduled and prioritized. All tiles in the footprint must be assigned to exactly one group.

Group names are arbitrary strings. The tiles associated with a group can be optionally restricted using:

• cap = N or S.
• dec_min or dec_max limits.
• covers keyword specifies previously defined subgroups that the tiles in this group cover (for fiber assignment).

Groups are also associated with a list of passes, with each pass defining a subgroup.

Each subgroup has associated rules that specify its relative weight when selecting the next tile in the scheduler.

Rules are assigned separately for each subgroup using the notation:

GROUP_NAME(PASS): { ...rules... }

By default, each subgroup has an initial weight of zero when the survey starts (so will not be scheduled), unless it has an explicit start rule:

START: INITIAL_WEIGHT

Additional rules of the form:

GROUP_NAME(PASS): NEW_WEIGHT

specify the subgroup weight that will be assigned when the specified subgroup has been completely observed. Forward references to subgroups that have not been defined yet are ok.