caf_top.py

 
 
"""
Airflow script for TOP post-tracking calibration:
   BS13d carrier shifts, module T0 and common T0.
   Histograms for validation are also prepared here.
"""
import basf2
from prompt import CalibrationSettings, INPUT_DATA_FILTERS
from caf.utils import IoV
from caf.strategies import SequentialBoundaries
from top_calibration import BS13d_calibration_cdst
from top_calibration import moduleT0_calibration_DeltaT, moduleT0_calibration_LL
from top_calibration import commonT0_calibration_BF
from top_calibration import offset_calibration
from top_calibration import calibration_validation
from prompt.calibrations.caf_top_pre import settings as top_pretracking
from prompt.utils import filter_by_max_files_per_run
 
 
settings = CalibrationSettings(
    name="TOP post-tracking calibration",
    expert_username="skohani",
    description=__doc__,
    input_data_formats=["cdst"],
    input_data_names=["mumu_tight_or_highm_calib"],
    input_data_filters={
        "mumu_tight_or_highm_calib": [
            INPUT_DATA_FILTERS["Data Tag"]["mumu_tight_or_highm_calib"],
            INPUT_DATA_FILTERS["Run Type"]["physics"],
            INPUT_DATA_FILTERS["Data Quality Tag"]["Good Or Recoverable"]]},
    depends_on=[top_pretracking],
    expert_config={
        "max_files_per_run": 20,
        "payload_boundaries": None,
        "request_memory": "8 GB"
    })
 
 
# Required function
def get_calibrations(input_data, **kwargs):
    '''
    Returns a list of calibration objects.
    :input_data (dict): Contains every file name from the 'input_data_names' as a key.
    :**kwargs: Configuration options to be sent in.
    '''
 
    file_to_iov = input_data["mumu_tight_or_highm_calib"]
    sample = 'dimuon'
    requested_iov = kwargs.get("requested_iov", None)
    expert_config = kwargs.get("expert_config")
    max_files_per_run = expert_config["max_files_per_run"]
    min_events_per_file = 1
    # Applying the min event per file to remove empty root files
    reduced_file_to_iov = filter_by_max_files_per_run(file_to_iov, max_files_per_run, min_events_per_file, random_select=True)
    inputFiles = list(reduced_file_to_iov.keys())
    basf2.B2INFO(f"Total number of files actually used as input = {len(inputFiles)}")
    requested_iov = kwargs.get("requested_iov", None)
    output_iov = IoV(requested_iov.exp_low, requested_iov.run_low, -1, -1)
 
    cal = [BS13d_calibration_cdst(inputFiles),  # this is run-dep
           moduleT0_calibration_DeltaT(inputFiles),  # this cal cannot span across experiments
           moduleT0_calibration_LL(inputFiles, sample),  # this cal cannot span across experiments
           commonT0_calibration_BF(inputFiles),  # this is run-dep
           offset_calibration(inputFiles),  # this is run-dep
           calibration_validation(inputFiles, sample)]  # this is run-dep
 
    for c in cal:
        # If it's a SequentialBoundary calibration, check if there is any boundary in the config file
        if c.strategies[0] == SequentialBoundaries:
 
            # Default boundaries. If there are no boundaries in the config file, this calibration will give a single IoV
            payload_boundaries = [[output_iov.exp_low, output_iov.run_low]]
 
            # user-defined boundaries are set here.
            if expert_config["payload_boundaries"] is not None:
                payload_boundaries = expert_config["payload_boundaries"]
 
            # Set the actual boundaries.
            for alg in c.algorithms:
                alg.params = {"iov_coverage": output_iov, "payload_boundaries": payload_boundaries}
 
        # If it's not a SequentialBoundary calbration, just set the IoV coverage
        else:
            for alg in c.algorithms:
                alg.params = {"iov_coverage": output_iov}
 
    # Don't save the rough moduleT0 result
    cal[1].save_payloads = False
    # Just to make it sure ...
    cal[5].save_payloads = False  # in fact it does not make any payloads, but produces histograms for validation
 
    cal[1].depends_on(cal[0])
    cal[2].depends_on(cal[1])
    cal[3].depends_on(cal[2])
    cal[4].depends_on(cal[3])
    cal[5].depends_on(cal[4])
 
    return cal