Utility functions and methods for evoland_db — evoland_db

Factored out utility functions and methods specifically used in conjunction with evoland_db and its subclasses.

Usage

get_evoland_db_read_expr(self, super, table_name)

run_parallel_evoland(items, worker_fun, parent_db, cluster = NULL, ...)

self: The evoland_db instance
table_name: The name of the table to read from
items: A list of items to iterate over
worker_fun: A function to apply to each item, with signature worker_fun(item, db, ...). The db argument will be a new evoland_db instance for each worker.
parent_db: An evoland_db instance; used for its path and id_run; or passed directly to the worker for serial case.
cluster: An optional cluster object created by parallel::makeCluster() or mirai::make_cluster().
...: Additional arguments passed to worker_fun.

A SQL expression string to read most specific data for slice

A list of results

get_evoland_db_read_expr(): Gets the read expression for a given table, taking into account the run lineage and distinctness columns. If the table has no id_run column or is the runs_t table itself, falls back to the default read expression. Otherwise, constructs a read expression that filters for the active id_run and its lineage, using the distinctness columns to determine which slices of data to include.
run_parallel_evoland(): Run a worker function in parallel with an evoland_db instance; takes an (optional) worker cluster, a worker function, a set of items, and a parent evoland_db instance. If no cluster is passed, apply the worker serially. If a cluster is passed, each worker gets its own database instance derived from the parent database.