CIME.baselines package
Submodules
CIME.baselines.performance module
- CIME.baselines.performance.get_cpl_mem_usage(cpllog)[source]
Read memory usage from coupler log.
Parameters
- cpllogstr
Path to the coupler log.
Returns
- list
Memory usage (data, highwater) as recorded by the coupler or empty list.
- CIME.baselines.performance.get_cpl_throughput(cpllog)[source]
Reads throuhgput from coupler log.
Parameters
- cpllogstr
Path to the coupler log.
Returns
- int or None
Throughput as recorded by the coupler or None
- CIME.baselines.performance.get_latest_cpl_logs(case)[source]
find and return the latest cpl log file in the run directory
- CIME.baselines.performance.load_coupler_customization(case)[source]
Loads customizations from the coupler cime_config directory.
Parameters
- caseCIME.case.case.Case
Current case object.
Returns
- CIME.config.Config
Runtime configuration.
- CIME.baselines.performance.perf_compare_memory_baseline(case, baseline_dir=None)[source]
Compares model highwater memory usage.
Parameters
- caseCIME.case.case.Case
Current case object.
- baseline_dirstr
Overrides the baseline directory.
Returns
- below_tolerancebool
Whether the comparison was below the tolerance.
- commentstr
Provides explanation from comparison.
- CIME.baselines.performance.perf_compare_throughput_baseline(case, baseline_dir=None)[source]
Compares model throughput.
Parameters
- caseCIME.case.case.Case
Current case object.
- baseline_dirstr
Overrides the baseline directory.
Returns
- below_tolerancebool
Whether the comparison was below the tolerance.
- commentstr
Provides explanation from comparison.
- CIME.baselines.performance.perf_get_memory(case, config)[source]
Gets the model memory usage.
First attempts to use a coupler defined method to retrieve the models memory usage. If this is not defined then the default method of parsing the coupler log is used.
Parameters
- caseCIME.case.case.Case
Current case object.
Returns
- str or None
Model memory usage.
- CIME.baselines.performance.perf_get_throughput(case, config)[source]
Gets the model throughput.
First attempts to use a coupler define method to retrieve the models throughput. If this is not defined then the default method of parsing the coupler log is used.
Parameters
- caseCIME.case.case.Case
Current case object.
Returns
- str or None
Model throughput.
- CIME.baselines.performance.perf_write_baseline(case, basegen_dir, throughput=True, memory=True)[source]
Writes the baseline performance files.
Parameters
- caseCIME.case.case.Case
Current case object.
- basegen_dirstr
Path to baseline directory.
- throughputbool
If true, write throughput baseline.
- memorybool
If true, write memory baseline.