Flow: Flow
In mskilab/Flow: Workflow and task management for genomics pipelines.

Flow	R Documentation

Flow

Description

Instantiates a Flow object which is a purely diagnostic / visualization utility to monitor progress across a set of entities x tasks.

Note: For job monitoring, it does not do anything fancy - just checks the paths pointed to by the columns of an entities table against task inputs and outputs to determine whether a task is ready or done.

However if some other task (or manual setting) of that column points to a valid path, then that pipeline will be marked as "done".

All job control should be done at the Job level. TBD: linking this object to a set of Flow object and allowing some job control at the Flow level.

Returns Flow object which then can be plotted or subsetted via [.Flow operator.

Usage

Flow(
  flowdir = "./Flow",
  taskdir = NULL,
  entities = NULL,
  update = FALSE,
  quick = FALSE,
  mc.cores = 1,
  paths.only = TRUE
)

Arguments

`flowdir`	path to Flow directory containing .rds of Job files, and/or paths to those .rds files
`taskdir`	path to task directory or set of task paths
`entities`	keyed entities data.table can be optionally provided with taskdir
`update`	logical flag whether to update jobs if flowdir provided (FALSE)
`quick`	won't check existence of file paths, just will check to see if non empty
`mc.cores`	integer multicore param
`paths.only`	whether to build the DAG only using path annotations rather than all arguments (otherwise too crowded)