dataset.DatasetBase Class Reference

Inheritance diagram for dataset.DatasetBase:

Public Member Functions
def	getfiles (self, usecache)
def	headercomment (self)
def	writefilelist_hippy (self, firstrun, lastrun, runs, eventsperjob, maxevents, outputfile, usecache=True)
def	writefilelist_validation (self, firstrun, lastrun, runs, maxevents, outputfile=None, usecache=True)

Static Private Attributes
	__metaclass__ = abc.ABCMeta

Detailed Description

Definition at line 127 of file dataset.py.

Member Function Documentation

def dataset.DatasetBase.getfiles	(		self,
			usecache
	)

Definition at line 131 of file dataset.py.

Referenced by dataset.DatasetBase.writefilelist_hippy().

  def getfiles(self, usecache):
    pass

def dataset.DatasetBase.headercomment

(

self

)

Definition at line 135 of file dataset.py.

  def headercomment(self):
    pass

def dataset.DatasetBase.writefilelist_hippy	(		self,
			firstrun,
			lastrun,
			runs,
			eventsperjob,
			maxevents,
			outputfile,
			usecache = True
	)

Definition at line 167 of file dataset.py.

References Vispa.Plugins.EdmBrowser.EdmDataAccessor.all(), any(), objects.autophobj.float, and dataset.DatasetBase.getfiles().

  def writefilelist_hippy(self, firstrun, lastrun, runs, eventsperjob, maxevents, outputfile, usecache=True):
    runrange = RunRange(firstrun=firstrun, lastrun=lastrun, runs=runs)
    if maxevents < 0: maxevents = float("inf")
    totalevents = sum(datafile.nevents for datafile in self.getfiles(usecache) if all(run in runrange for run in datafile.runs))
    if totalevents == 0:
      raise ValueError("No events within the run range!")
    accepted = rejected = inthisjob = 0.  #float so fractions are easier

    fractiontoaccept = 1.*maxevents / totalevents
    writecomma = False

    with open(outputfile, "w") as f:
      for datafile in self.getfiles(usecache):
        if all(run in runrange for run in datafile.runs):
          if accepted == 0 or accepted / (accepted+rejected) <= fractiontoaccept:
            if writecomma: f.write(",")
            f.write("'" + datafile.filename + "'")
            accepted += datafile.nevents
            inthisjob += datafile.nevents
            if inthisjob >= eventsperjob:
              f.write("\n")
              inthisjob = 0
              writecomma = False
            else:
              writecomma = True
          else:
            rejected += datafile.nevents
        elif any(run in runrange for run in datafile.runs):
          raise DatasetError("file {} has multiple runs {}, which straddle firstrun or lastrun".format(datafile.filename, datafile.runs))
      f.write("\n")

def dataset.DatasetBase.writefilelist_validation	(		self,
			firstrun,
			lastrun,
			runs,
			maxevents,
			outputfile = None,
			usecache = True
	)

Definition at line 138 of file dataset.py.

  def writefilelist_validation(self, firstrun, lastrun, runs, maxevents, outputfile=None, usecache=True):
    runrange = RunRange(firstrun=firstrun, lastrun=lastrun, runs=runs)

    if outputfile is None:
      outputfile = os.path.join(os.environ["CMSSW_BASE"], "src", "Alignment", "OfflineValidation", "python", self.filenamebase+"_cff.py")

    if maxevents < 0: maxevents = float("inf")
    totalevents = sum(datafile.nevents for datafile in self.getfiles(usecache) if all(run in runrange for run in datafile.runs))
    if totalevents == 0:
      raise ValueError("No events within the run range!")
    accepted = rejected = 0.  #float so fractions are easier

    fractiontoaccept = 1.*maxevents / totalevents

    with open(outputfile, "w") as f:
      f.write("#"+self.headercomment+"\n")
      f.write(validationheader)
      for datafile in self.getfiles(usecache):
        if all(run in runrange for run in datafile.runs):
          if accepted == 0 or accepted / (accepted+rejected) <= fractiontoaccept:
            f.write('"' + datafile.filename + '",\n')
            accepted += datafile.nevents
          else:
            rejected += datafile.nevents
        elif any(run in runrange for run in datafile.runs):
          raise DatasetError("file {} has multiple runs {}, which straddle firstrun or lastrun".format(datafile.filename, datafile.runs))
      f.write("#total events in these files: {}".format(accepted))
      f.write(validationfooter)

Member Data Documentation

dataset.DatasetBase.__metaclass__ = abc.ABCMeta

staticprivate

Definition at line 128 of file dataset.py.