def main(input, algorithmAttributes):
assert input, 'no input is provided, the algorithm cannot run'
ImportInput(input, algorithmAttributes)
AllocManagement_Hybrid2()
outputResults()
def main(input, algorithmAttributes):
startTime = time.time()
ver = ImportInput(input, algorithmAttributes)
if ver == "stop":
return
if G.ACO == "all":
G.acoRange = [0,1]
G.minRange = {0:[0,1],1:[0,1]}
elif G.ACO == "1":
G.acoRange = [1]
G.minRange = {0:[0,1], 1:[G.minDeltaUt]}
else:
G.acoRange = [0]
G.minRange = {0:[G.minDeltaUt]}
for j in G.acoRange:
for i in G.minRange[j]:
initialiseVar()
G.minDeltaUt = i
G.ACO = j
print 'start ACO', G.ACO, 'minDelta', G.minDeltaUt
bestAnt = AllocManagement_Hybrid2(None)
# salvare risultati
G.Summary[(G.ACO,G.minDeltaUt)] = {'scenario':(G.ACO,G.minDeltaUt)}
for key in G.LateMeasures.keys():
if key == 'lateness' or key == 'earliness':
if len(G.LateMeasures[key]):
G.Summary[(G.ACO,G.minDeltaUt)][key] = mean(G.LateMeasures[key])
else:
G.Summary[(G.ACO,G.minDeltaUt)][key] = 0
else:
G.Summary[(G.ACO,G.minDeltaUt)][key] = G.LateMeasures[key]
utilisation = []
targetUt = []
for bottleneck in G.Bottlenecks:
for week in G.WeekList:
utilisation.append(float(G.Capacity[bottleneck][week]['OriginalCapacity']-G.CurrentCapacityDict[bottleneck][week])/G.Capacity[bottleneck][week]['OriginalCapacity'])
if G.Capacity[bottleneck][week]['targetUtilisation']:
targetUt.append((G.Capacity[bottleneck][week]['targetUtilisation']-float(G.Capacity[bottleneck][week]['OriginalCapacity']-G.CurrentCapacityDict[bottleneck][week])/G.Capacity[bottleneck][week]['OriginalCapacity'])/G.Capacity[bottleneck][week]['targetUtilisation'])
else:
targetUt.append((G.Capacity[bottleneck][week]['targetUtilisation']-float(G.Capacity[bottleneck][week]['OriginalCapacity']-G.CurrentCapacityDict[bottleneck][week])/G.Capacity[bottleneck][week]['OriginalCapacity']))
G.Summary[(G.ACO,G.minDeltaUt)]['utilisation'] = mean(array(utilisation))
G.Summary[(G.ACO,G.minDeltaUt)]['targetM'] = mean(absolute(array(targetUt)))
G.Summary[(G.ACO,G.minDeltaUt)]['targetStd'] = std(array(targetUt))
if G.ACO:
G.Summary[(G.ACO,G.minDeltaUt)]['ant'] = bestAnt
else:
G.Summary[(G.ACO,G.minDeltaUt)]['ant'] = None
# selection
listSummary = [G.Summary[item] for item in G.Summary.keys()]
print 'list summary', listSummary
listSummary = sorted(listSummary, key=itemgetter('exUnits', 'lateness', 'targetM', 'targetStd', 'utilisation', 'earliness'))
bestScenario = listSummary[0]['scenario']
aco = bestScenario[0]
minDelta = bestScenario[1]
G.Summary['orderedScenario'] = {}
for i in range(len(listSummary)):
G.Summary['orderedScenario'][listSummary[i]['scenario']] = i+1
G.Summary['bestScenario'] = bestScenario
if aco != G.ACO or minDelta != G.minDeltaUt:
initialiseVar()
G.ACO = 0 # forces the simulation of the best ant (even though the best scenario is ACO
G.minDeltaUt = minDelta
AllocManagement_Hybrid2(G.Summary[(aco,minDelta)]['ant'])
AllocManagement_Hybrid2_Forecast()
outputResults()
print 'calculation time', time.time()-startTime
def main(input):
assert input, 'no input is provided, the algorithm cannot run'
ImportInput(input)
AllocManagement_Hybrid2()
outputResults()
