sciline.reporter.TimingReporter#

class sciline.reporter.TimingReporter[source]#

A reporter that tracks the time spent in each provider.

Examples

from sciline.reporter import TimingReporter
timer = TimingReporter()
pipeline.compute(Target, reporter=timer)
print(timer.summary())

Methods

`__init__`()
`as_pandas`()	Return the recorded times as a Pandas DataFrame.
`call_provider_with_reporting`(provider, values)	Call a provider and report its progress with this reporter.
`on_provider_end`(provider_id)	Stop the clock for a provider call.
`on_provider_start`(provider)	Start the clock for a provider call.
`reporting_provider_func`(provider)	Wrap a provider's function to report progress with this reporter.
`run_computation`(providers)	Run a single computation with progress reporting.
`summary`()	Return a summary of the recorded times.

as_pandas()[source]#

Return the recorded times as a Pandas DataFrame. :rtype: DataFrame

See also

TimingReporter.summary: Return a summary of the recorded times as a string. Does not require Pandas.

call_provider_with_reporting(provider, values)#

Call a provider and report its progress with this reporter.

on_provider_end(provider_id)[source]#

Stop the clock for a provider call.

on_provider_start(provider)[source]#

Start the clock for a provider call.

reporting_provider_func(provider)#

Wrap a provider’s function to report progress with this reporter.

run_computation(providers)#

Run a single computation with progress reporting.

This method is primarily meant for internal use. As a user, you should pass reporters to Pipeline.compute().

Use this method in a with statement:

with reporter.run_computation(providers):
    # Perform computation

This method exists mainly to provide arguments when entering a reporting context, implementers should mainly override __enter__.

Return a summary of the recorded times.