cosapp.systems.batch

Functions

batch_run(system, inputs, output_varnames[, ...])

Convenience function to run a batch execution on a system.

Classes

BatchRunner(system, output_varnames[, policy])

Batch execution framework for systems.
class cosapp.systems.batch.BatchRunner(system: System, output_varnames: Sequence[str], policy: ExecutionPolicy | None = None)[source]

Bases: object

Batch execution framework for systems.

Parameters:

  • [System] (- system) – The system of interest.

  • [Sequence[str]] (- output_varnames) – Variable names to be captured as inputs.

  • [ExecutionPolicy] (- policiy) – Execution policy for batch runs.

static empty_dataset(varnames: Sequence[str]) → dict[str, list][source]

Generate a dictionary mapping each variable name to an empty list.

property execution_policy: ExecutionPolicy

Default execution policy of the batch runner.

static filter_headers(dataframe: DataFrame, inplace=True) → DataFrame[source]

Apply natural_varname to dataframe column names

find_outputs(includes: str | list[str] | type[None] = '*', excludes: str | list[str] | type[None] = []) → None[source]

Set output variable names matching inclusion and exclusion patterns

classmethod from_output_pattern(system: System, includes: SearchPattern, excludes: SearchPattern, policy: ExecutionPolicy | None = None) → BatchRunner[source]

Factory constructing a BatchRunner instance from specific search patterns for output variables.

property output_varnames

output variable names captured by runner

Type:

“tuple[str]

run(inputs: DataFrame, policy: ExecutionPolicy | None = None) → dict[str, list][source]

Batch execution.

Parameters:

  • [pandas.DataFrame] (- inputs) – Dataframe containing batches of input values, referenced by variable names.

  • [ExecutionPolicy (- policy) – Execution policy for the job. If not prescribed (default), uses the runner policy set at initialization.

  • optional] – Execution policy for the job. If not prescribed (default), uses the runner policy set at initialization.

Returns:

dict[str, list]

Return type:

output data, stored as a dictionary of the kind {varname, list[values]}.

property system: System

System executed by the runner

cosapp.systems.batch.batch_run(system: System, inputs: pandas.DataFrame, output_varnames: Sequence[str], execution_policy: ExecutionPolicy | None = None) → dict[str, list][source]

Convenience function to run a batch execution on a system.

Parameters:

  • [System] (- system)

  • [pandas.DataFrame] (- inputs)

  • [Sequence[str]] (- output_varnames)

  • [Optional[ExecutionPolicy]] (- execution_policy) – If not provided, defaults to single-thread execution.

Returns:

dict[str, list]

Return type:

Output data as a dictionary of lists.