planner

Summary:

Cached state of the survey afternoon planner.

Naming Convention:

planner_YEAR-MM-DD.fits, where YEAR-MM-DD is the date of the afternoon when the planner was run.

Regex:

planner_[0-9]{4}-[0-9]{2}-[0-9]{2}\.fits

File Type:

FITS, 400 KB

Contents

Number	EXTNAME	Type	Contents
HDU0		IMAGE	Blank.
HDU1	PLAN	BINTABLE	Survey planner state.

FITS Header Units

HDU0

EXTNAME = (None)

Empty HDU.

HDU1

EXTNAME = PLAN

Snapshot of the internal state of a desisurvey.plan.Planner object.

Required Header Keywords

KEY	Example Value	Type	Comment
NAXIS1	29	int	length of dimension 1.
NAXIS2	16071	int	length of dimension 2.
CADENCE	monthly	str	Fiber assignment cadence (monthly / daily).
FIRST	2019-12-01	str	First night that was planned.
LAST	2024-11-29	str	Last night that has been planned so far.

Required Data Table Columns

Name	Type	Units	Description
TILEID	int64		Tile ID from the tiles file.
COVERED	int64		Day number that tile was first covered relative to FIRST or -1.
COUNTDOWN	int64		Countdown of remaining fiber assignment cycles until this tile will be assigned.
AVAILABLE	logical		Does this tile have fibers already assigned?
PRIORITY	float64		Relative priority for scheduling this tile. Must be >= 0.

The meaning of “covered” is specified by the fiber_assignment_order configuration parameter, which also specifies the starting COUNTDOWN value for each pass.

Each row of the table corresponds to one tile with indexing that matches desisurvey.tiles.Tiles.

Notes and Examples

A Planner object manages updates during afternoon planning:

import desisurvey.rules
import desisurvey.plan
rules = desisurvey.rules.Rules()
planner = desisurvey.plan.Planner(rules)

Its internal state after each afternoon can be saved using, for example:

planner.save('planner_snapshot.fits')

This state can then be later restored using:

planner = desisurvey.plan.Planner(rules, restore='planner_snapshot.fits')