![]() |
ATLAS Offline Software
|
Classes | |
| class | taskIterator |
Functions | |
| getRunLumiInfoFragment (jobnumber, task, maxEvents, totalEvents, skipEvents, sequentialEventNumbers=False) | |
| getRandomlySampledRunLumiInfoFragment (jobnumber, task, maxEvents, totalEvents, skipEvents, sequentialEventNumbers=False) | |
| getFragment (jobnumber, task, maxEvents) | |
| findPlaceInTask (jobnumber, task, maxEvents) | |
| taskLookupTable (task) | |
| defineSequentialEventNumbers (jobnumber, fragment, totalEvents, skipEvents) | |
| python.RunDependentMCTaskIterator.defineSequentialEventNumbers | ( | jobnumber, | |
| fragment, | |||
| totalEvents, | |||
| skipEvents ) |
Calculate sequential event numbers for the defined getFragment.
Definition at line 164 of file RunDependentMCTaskIterator.py.
| python.RunDependentMCTaskIterator.findPlaceInTask | ( | jobnumber, | |
| task, | |||
| maxEvents ) |
Get the 'i'th job in the task, where each job tries to do maxEvents events. The 'force_new' flag in the LB list ends the task before that LB, ignoring maxEvents. Returns a taskIterator. Can raise StopIteration, so you should nest in try.
Definition at line 91 of file RunDependentMCTaskIterator.py.
| python.RunDependentMCTaskIterator.getFragment | ( | jobnumber, | |
| task, | |||
| maxEvents ) |
Calculate the specific configuration of the current job in the digi task.
Definition at line 80 of file RunDependentMCTaskIterator.py.
| python.RunDependentMCTaskIterator.getRandomlySampledRunLumiInfoFragment | ( | jobnumber, | |
| task, | |||
| maxEvents, | |||
| totalEvents, | |||
| skipEvents, | |||
| sequentialEventNumbers = False ) |
Calculate the specific configuration of the current job in the digi task. Sample the mu values randomly.
Definition at line 42 of file RunDependentMCTaskIterator.py.
| python.RunDependentMCTaskIterator.getRunLumiInfoFragment | ( | jobnumber, | |
| task, | |||
| maxEvents, | |||
| totalEvents, | |||
| skipEvents, | |||
| sequentialEventNumbers = False ) |
Calculate the specific configuration of the current job in the digi task. Try to make each fragment utilize the same amount of CPU and Cache resources. Exploits the fact that the task when sorted by mu value will have long finishing pieces near the beginning and short pieces near the end. A more even solution is obtained by pairing chunks of maxEvents/2 from the beginning and end of the sorted task.
Definition at line 14 of file RunDependentMCTaskIterator.py.
| python.RunDependentMCTaskIterator.taskLookupTable | ( | task | ) |
Generate task lookup table
Definition at line 155 of file RunDependentMCTaskIterator.py.