processPool=processPool,
featureName=settingsDict['Feature Name'],
numFeatures=int(settingsDict['Number of Features']),
downSampleFactor=int(settingsDict['Feature DownSample Factor']),
timeDelay=int(settingsDict['CRP Time Delay']),
dimension=int(settingsDict['CRP Dimension']),
method=settingsDict['CRP Method'],
neighbourhoodSize=settingsDict['CRP Neighbourhood Size'],
runType='validation',
sequenceLength=int(settingsDict['NCD Sequence Length']),
weightMatrix=weightMatrix,
biases=biases,
featureOffset=featureOffset,
featureScaling=featureScaling,
timeStacking=int(settingsDict['NN Time Stacking']))
# Convert NCD files into a dataframe
dfNCDs = convertNCDfiles(dataFrameFileName=settingsDict['Run Name'])
# Get the overall MAP of the run and add to the setting
MAPresult = getDataFrameMAPresult(dfNCDs)
if MAPresult is not None:
print 'Mean Average Precision: %0.3f\n' % MAPresult
else:
print 'No MAP result found!'
settingsDict['Mean Average Precision'] = MAPresult
settingsDict['Run Duration'] = datetime.now() - startDateTime
# Save run settings and result
pickle.dump(settingsDict, open(resultsFn, 'wb'))
featureName, setting['DownSample Factor'],
setting['Time Delay'], setting['Dimension'], CRPmethod, setting['Neighbourhood Size'],
numFilesPerFolder, setting['Sequence Length'],
weightMatrix, biases, featureOffset, featureScaling)
# create and save a record of the run settings
runDict = {'featureName': featureName,
'method': CRPmethod,
'numFilesPerFolder': numFilesPerFolder,
'networkType': NNtype}
for key in setting.keys():
runDict[key] = setting[key]
runTime = str(datetime.now()).replace(':', '-')
pickle.dump(runDict, open(runHistoryPath + runTime + '.pkl', 'wb'))
# Convert NCD files into a dataframe
dfNCDs = convertNCDfiles(dataFrameFileName = runTime)
# Create subfolders and move results files into them
NCDdest = NCDpath + subFolder + '/'
runHistDest = runHistoryPath + subFolder + '/'
createPath(NCDdest)
createPath(runHistDest)
moveFiles(NCDpath, NCDdest, '.pkl.res')
moveFiles(runHistoryPath, runHistDest, '.pkl')
# Get the overall MAP of the run and add to the setting
MAPresult = getDataFrameMAPresult(dfNCDs)
if MAPresult is not None:
print 'Mean Average Precision: %0.3f\n' % MAPresult
else:
print 'No MAP result found!'
deleteNCDPickleFiles()
existingNCDs = None
processPool = Pool(numProcesses)
for downSampleFactor in downSampleFactors[::-1]:
for neighbourhoodSize in CRPneighbourhoodSizes:
for method in CRPmethods:
for dimension in CRPdimensions:
for timeDelay in CRPtimeDelays:
for sequenceLength in sequenceLengths:
createNCDfiles(existingNCDs, processPool,
featureName, downSampleFactor,
timeDelay, dimension, method, neighbourhoodSize,
numFilesPerFolder, sequenceLength)
runDict = {'featureName': featureName,
'downSampleFactor': downSampleFactor,
'timeDelay': timeDelay,
'dimension': dimension,
'method': method,
'neighbourhoodSize': neighbourhoodSize,
'numFilesPerFolder': numFilesPerFolder,
'sequenceLength': sequenceLength}
runTime = str(datetime.now()).replace(':', '-')
pickle.dump(runDict, open(runHistoryPath + runTime + '.pkl', 'wb'))
convertNCDfiles(runTime)