trplearn.scheduling module#

class trplearn.scheduling.PERTAnalyst[source]#

Bases: object

Project Evaluation and Review Technique (PERT) analyst using Max-Plus algebra. This class computes earliest/latest times and critical paths by interpreting the Kleene star (closure) of a task duration matrix.

The input adjacency matrix A should have A[i, j] = duration of task from event i to j. Use -np.inf where no direct task exists between events.

longest_path_matrix_#

The all-pairs longest path matrix (Kleene star).

Type:: MaxPlus

durations_#

The original task duration matrix.

Type:: ndarray

compute(adjacency_matrix)[source]#

Compute the longest path matrix and prepare for analysis.

Parameters:: adjacency_matrix (array-like) – Square matrix where [i, j] is the duration of the task starting at event i and ending at event j. Use -np.inf for no dependency.
Returns:: self
Return type:: object

critical_path()[source]#

Identify the sequence of event indices forming the critical path. An event is on a critical path if its slack is zero.

Returns:: List of event indices on the critical path.
Return type:: list of int

property earliest_times#

Earliest Start Times (EST) for each event, assuming project starts at event 0.

Return type:: ndarray

property latest_times#

Latest Start Times (LST) for each event without delaying the project.

Return type:: ndarray

project_duration()[source]#

Total duration of the project (earliest time of the final event).

Return type:: float

property slack#

Total slack (float) for each event. Slack = Latest Time - Earliest Time.

Return type:: ndarray