Schedules

Wiki / Explorer

The Schedule view

Select Schedule to see a result as a Gantt chart. Explorer opens it by itself when a result holds the schedule map from gdb.schedule.solve.stats:

MATCH (d:Dance) WITH collect(d) AS dances
MATCH (s:Stage) WITH dances, collect(s) AS stages
CALL gdb.schedule.solve.stats({tasks: dances, resources: stages,
  task: {participants: '[(p:Dancer)-[:DANCES_IN]->(t) | p]'},
  participant: {changeover: 15}})
YIELD schedule RETURN schedule;

Each timeline is a lane. Lanes switches between resources (the stages the scheduler chose), participants (each dancer's day) and both. A task that occupies several timelines, such as a dance with four dancers or an operation with a surgeon, an anaesthetist and a theatre, appears on each of them; click a bar to see its times and every timeline it uses, and the rest of the chart dims. Time outside a timeline's windows is shaded, a timeline that holds several tasks at once (units) shares its lane in stripes, and the lane's share of its open time is shown beside its name. Unassigned tasks are grouped by reason, and the objective is listed tier by tier.

Rows work too: any result with a start and an end per row, such as the rows of gdb.schedule.solve or bookings you wrote back, is drawn with one lane per resource or resources value. Times can be numbers or any temporal type; date-times are shown in their own offset.

Drag to pan, scroll or use + / − to zoom time, and use Fit to show the whole plan. The focused chart also takes the left and right arrow keys, + / −, 0 (fit) and Escape (clear the selection). The view draws up to 20,000 items on 2,000 timelines.

The schedule builder

▤ Schedule in the editor toolbar opens a builder that writes the gdb.schedule.solve call for you. Choose the task and resource labels, the duration and preferred start expressions, participants through a relationship, when resources are available, who can take what (any resource, a skills match through two relationship types, or your own expression), rules such as most work or most tasks per day, and the order of the objective tiers. Label, relationship and property names are suggested from the database's schema.

The Cypher updates as you edit. Insert into editor puts it in the current tab; Run also runs it, as a chart preview (solve.stats), one row per task (solve), or by writing a booking relationship for each placement. The builder remembers its last settings in this browser. Everything it writes is ordinary Cypher, so you can carry on from there in the editor.

Scheduling · Write and run queries · Graph data science

Planning a deployment? Review compatibility and licence setup for your instance.