def AllocationRoutine_ForecastGA(initialWeek, itemList, itemType, chromo):

    EarlinessMA = {}
    LatenessMA = {}

    GAexcess = 0
    GAcapacityDict = deepcopy(G.CurrentCapacityDict)
    GAincompleteBatches = deepcopy(G.incompleteBatches)
    GAearliness = 0
    GAlateness = 0
    GAtargetUtil = 0
    GAminUtil = 0
    
    # repeat allocation procedure for all items in the list
    for order in chromo['seq']:
        
        item=itemList[order]
        
#        print 'item', item['orderID']
        
        #================================================
        # Allocation step 1...allocation at current Week
        #================================================
        
        # variables created for one specific order
        # will be confirmed in the GA variables provided that there is enough capacity to allocate the entire order (possibly across different weeks)
        Results = {}
        step = 1
        ind = G.WeekList.index(initialWeek)
        weekList = [initialWeek]
        capacity = deepcopy(GAcapacityDict)
        inBatches = deepcopy(GAincompleteBatches)
        qty = item['Qty']
        Allocation = []
        earliness = 0
        lateness = 0
        previousAss = {}
        for ma in G.SPlist[item['sp']]:
            previousAss[ma] = 0

        
        while step <= 3 and qty>0:
            
            if step == 2:                    
                weekList = [G.WeekList[i] for i in range(ind-1, max(-1,ind-G.maxEarliness-1), -1)]
                
            if step == 3:
                weekList = [G.WeekList[i] for i in range(ind+1, min(G.planningHorizon,ind+G.maxLateness+1))]
            
#            print 'weeklist', weekList
            if len(weekList) == 0:
                step+=1
                continue
            
            # allocate requested qty 
            for week in weekList:
                
                # optimise MA allocation at current week        
                spAllocation = Allocation_IP(item, week, previousAss, capacity, G.weightFactor)

                # implement optimal MA solution to update temporary variables
                for ma in spAllocation.keys():
                    if spAllocation[ma]:
                        Results = Allocation2(ma, spAllocation[ma], [week], capacity, inBatches, earliness, lateness, Allocation, initialWeek)
                        assert (Results['remainingUnits'] == 0)
                        
                        # update order variables
                        capacity = deepcopy(Results['remainingCap'])
                        inBatches = deepcopy(Results['remUnits'])
                        qty -= spAllocation[ma]
                        Allocation = deepcopy(Results['Allocation'])
                        earliness = deepcopy(Results['earliness'])
                        lateness = deepcopy(Results['lateness'])
                        if ma not in EarlinessMA:
                            EarlinessMA[ma] = 0
                            LatenessMA[ma] = 0
                        EarlinessMA[ma] += max([0, initialWeek - week])*spAllocation[ma]
                        LatenessMA[ma] += max([0, week - initialWeek])*spAllocation[ma]
                            
                        previousAss[ma] += spAllocation[ma]
                
                # if order has been fully allocated update GA variables        
                if qty <= 0:
                    GAcapacityDict = deepcopy(capacity)
                    GAincompleteBatches = inBatches
                    GAearliness += earliness/item['Qty']
                    GAlateness += lateness/item['Qty']                    
                    break
            
            step += 1
        
        # if order can not been confirmed then update the GA excess variable
        if qty > 0:
            GAexcess += item['Qty']
    
    if G.minDeltaUt:       
        GAtargetUtil, GAminUtil = utilisationCalc1(GAcapacityDict, initialWeek, ind)
    else:
        GAtargetUtil, GAminUtil = utilisationCalc2(GAcapacityDict, initialWeek, ind)
        
    return {'chromo':chromo, 'excess':GAexcess, 'earliness':GAearliness, 'lateness':GAlateness, 'targetUtil':GAtargetUtil, 'minUtil':GAminUtil}


        
def AllocationRoutine_ForecastGA(initialWeek, itemList, itemType, chromo):

    GAexcess = 0
    GAcapacityDict = deepcopy(G.CurrentCapacityDict)
    GAincompleteBatches = deepcopy(G.incompleteBatches)
    GAearliness = 0
    GAlateness = 0
    GAtargetUtil = 0
    GAminUtil = 0

    # repeat allocation procedure for all items in the list
    for order in chromo['seq']:

        item = itemList[order]

        print 'item', item['orderID']

        #================================================
        # Allocation step 1...allocation at current Week
        #================================================

        # variables created for one specific order
        # will be confirmed in the GA variables provided that there is enough capacity to allocate the entire order (possibly across different weeks)
        Results = {}
        step = 1
        ind = G.WeekList.index(initialWeek)
        weekList = [initialWeek]
        capacity = deepcopy(GAcapacityDict)
        inBatches = deepcopy(GAincompleteBatches)
        qty = item['Qty']
        Allocation = []
        earliness = 0
        lateness = 0
        previousAss = {}
        for ma in G.SPlist[item['sp']]:
            previousAss[ma] = 0

        while step <= 3 and qty > 0:

            if step == 2:
                weekList = [
                    G.WeekList[i]
                    for i in range(ind - 1, max(-1, ind - G.maxEarliness -
                                                1), -1)
                ]

            if step == 3:
                weekList = [
                    G.WeekList[i] for i in range(
                        ind + 1, min(G.planningHorizon, ind + G.maxLateness +
                                     1))
                ]

#            print 'weeklist', weekList
            if len(weekList) == 0:
                step += 1
                continue

            # allocate requested qty
            for week in weekList:

                # optimise MA allocation at current week
                spAllocation, probStatus = Allocation_IP(
                    item, week, previousAss, capacity, G.weightFactor)

                # implement optimal MA solution to update temporary variables
                for ma in spAllocation.keys():

                    if probStatus == 'Optimal':
                        allocatedQty = spAllocation[ma]
                    else:
                        allocatedQty = ceil(spAllocation[ma])

                    if spAllocation[ma]:

                        #print 'spAllocation', ma, spAllocation[ma], inBatches[ma]
                        Results = Allocation2(ma, allocatedQty, [week],
                                              capacity, inBatches, earliness,
                                              lateness, Allocation,
                                              initialWeek)
                        #print 'rem units', Results['remainingUnits']

                        if probStatus == 'Optimal':
                            assert (Results['remainingUnits'] == 0)
                        else:
                            allocatedQty -= Results['remainingUnits']

                        # update order variables
                        capacity = deepcopy(Results['remainingCap'])
                        inBatches = deepcopy(Results['remUnits'])
                        qty -= allocatedQty
                        Allocation = deepcopy(Results['Allocation'])
                        earliness = deepcopy(Results['earliness'])
                        lateness = deepcopy(Results['lateness'])

                        previousAss[ma] += allocatedQty

                # if order has been fully allocated update GA variables
                if qty <= 0:
                    GAcapacityDict = deepcopy(capacity)
                    GAincompleteBatches = inBatches
                    GAearliness += earliness / item['Qty']
                    GAlateness += lateness / item['Qty']
                    break

            step += 1

        # if order can not been confirmed then update the GA excess variable
        if qty > 0:
            GAexcess += item['Qty']

    if G.minDeltaUt:
        GAtargetUtil, GAminUtil = utilisationCalc1(GAcapacityDict, initialWeek,
                                                   ind)
    else:
        GAtargetUtil, GAminUtil = utilisationCalc2(GAcapacityDict, initialWeek,
                                                   ind)

    return {
        'chromo': chromo,
        'excess': GAexcess,
        'earliness': GAearliness,
        'lateness': GAlateness,
        'targetUtil': GAtargetUtil,
        'minUtil': GAminUtil
    }
def AllocationRoutine_Forecast(initialWeek, itemList, itemType, seq):
    
    
    
    # repeat allocation procedure for all items in the list
    for order in seq['seq']:
        
        item = itemList[order]
#        print 'item', item['orderID']
        
        EarlinessMA = {}
        LatenessMA = {}
        lateForecast = 0
        earlyForecast = 0
        #================================================
        # Allocation step 1...allocation at current Week
        #================================================
        
        Results = {}
        step = 1
        ind = G.WeekList.index(initialWeek)
        weekList = [initialWeek]
        weekLateness = [0]
        capacity = deepcopy(G.CurrentCapacityDict)
        incompleteBatches = deepcopy(G.incompleteBatches)
        qty = item['Qty']
        Allocation = []
        earliness = 0
        lateness = 0
        previousAss = {}
        for ma in G.SPlist[item['sp']]:
            previousAss[ma] = 0
            previousAss[ma] = incompleteBatches[ma]
#            print 'ma in', ma, incompleteBatches[ma]
            qty -= incompleteBatches[ma]
            # make use of incomplete batch units available
            incompleteBatches[ma] = 0
        
#        print 'order qty', qty   
        
        while step <= 3 and qty>0:
            
            if step == 2:                    
                weekList = [G.WeekList[i] for i in range(ind-1, max(-1,ind-G.maxEarliness-1), -1)]
                weekLateness = [i-ind for i in range(ind-1, max(-1,ind-G.maxEarliness-1), -1)]
                assert (len(weekList)==len(weekLateness))
                
            if step == 3:
                weekList = [G.WeekList[i] for i in range(ind+1, min(G.planningHorizon,ind+G.maxLateness+1))]
                weekLateness = [i-ind for i in range(ind+1, min(G.planningHorizon,ind+G.maxLateness+1))]
                
#            print 'weeklist', weekList
            if len(weekList) == 0:
                step+=1
                continue
            
            # check different MAs
            for weekIndex in range(len(weekList)):
                
                week = weekList[weekIndex]
                
                # optimise MA allocation               
                spAllocation, probStatus = Allocation_IP(item, week, previousAss, capacity,G.weightFactor)
#                print 'all', spAllocation
                # implement optimal MA solution
                for ma in spAllocation.keys():
                    if spAllocation[ma]:
                        Results = Allocation2(ma, spAllocation[ma], [week], capacity, incompleteBatches, earliness, lateness, Allocation, initialWeek)
                        assert (Results['remainingUnits'] == 0)
                        
                        # update variables
                        capacity = deepcopy(Results['remainingCap'])
                        qty -= spAllocation[ma]
                        Allocation = deepcopy(Results['Allocation'])
                        earliness = deepcopy(Results['earliness'])
                        lateness = deepcopy(Results['lateness'])
                        if ma not in EarlinessMA:
                            EarlinessMA[ma] = 0
                            LatenessMA[ma] = 0
                        
                        EarlinessMA[ma] += max([0, weekLateness[weekIndex]*(-1)])*spAllocation[ma]
                        LatenessMA[ma] += max([0, weekLateness[weekIndex]])*spAllocation[ma]     #week - initialWeek
                        
                        previousAss[ma] += spAllocation[ma]
                        
                if qty <= 0:
                    break
            
            step += 1
        
        # confirm results 
        if qty <= 0:
            
            minAss =item['Qty']*10
            maIB = None
            G.CurrentCapacityDict = Results['remainingCap']
            G.incompleteBatches = Results['remUnits']
#            print initialWeek, G.Earliness
            for maT in EarlinessMA:
#                assert(Results['remUnits'][maT]==0)
                G.Earliness[initialWeek][maT]['qty'].append(item['Qty'])
                G.Earliness[initialWeek][maT]['earliness'].append(EarlinessMA[maT]/item['Qty'])
                G.Lateness[initialWeek][maT]['qty'].append(item['Qty'])
                G.Lateness[initialWeek][maT]['lateness'].append(LatenessMA[maT]/item['Qty'])
                lateForecast += LatenessMA[maT]/item['Qty']
                earlyForecast += EarlinessMA[maT]/item['Qty']
                if previousAss[maT] <= minAss:
                    minAss = previousAss[maT]
                    maIB = maT
            
#            print 'ma ib', maIB, qty, cQty, item['Qty'], previousAss, EarlinessMA , [G.incompleteBatches[ma] for ma in G.SPlist[item['sp']]]
            if maIB != None:
                G.incompleteBatches[maIB] -= qty 
#            print [G.incompleteBatches[ma] for ma in G.SPlist[item['sp']]]
                
            G.orders[item['orderID']]['Allocation'] = Results['Allocation']
            G.orders[item['orderID']]['Excess'] = False
            chosenMA = []
            for allMA in Results['Allocation']:
                if allMA['ma'] not in chosenMA:
                    chosenMA.append(allMA['ma'])
            G.orders[item['orderID']]['chosenMA'] = chosenMA    
            
            if lateForecast:
                G.LateMeasures['noLateOrders'] += 1
                G.LateMeasures['lateness'].append(lateForecast)
                
            if earlyForecast:
                G.LateMeasures['noEarlyOrders'] += 1
                G.LateMeasures['earliness'].append(earlyForecast)            
                           
    
            for allRep in Results['Allocation']:
                G.globalMAAllocation[allRep['ma']][allRep['week']][itemType][item['priority']] += allRep['units']
                G.globalMAAllocationIW[allRep['ma']][initialWeek][itemType][item['priority']] += allRep['units']
            
        
            if itemType == 'forecast':
                G.forecastResults.append((item['ppos'], item['sp'], item['MAlist'], item['Week'], item['Qty'], item['priority'], chosenMA, Results['lateness']/item['Qty'], Results['earliness']/item['Qty'], Results['Allocation']))
                
        
        else:
            G.Excess[item['sp']][initialWeek] += item['Qty']
            G.orders[item['orderID']]['Allocation'] = []
            G.orders[item['orderID']]['Excess'] = True
            G.orders[item['orderID']]['chosenMA'] = None
            G.globalMAAllocationIW[item['sp']][initialWeek][itemType][item['priority']] += item['Qty']
            G.LateMeasures['noExcess'] += 1
            G.LateMeasures['exUnits'] += item['Qty']
            if itemType == 'forecast':
                G.forecastResults.append((item['ppos'], item['sp'], item['MAlist'], item['Week'], item['Qty'], item['priority'], None, 'NaN', 'NaN', 'None'))
def AllocationRoutine_Forecast(initialWeek, itemList, itemType, seq):
    
    
    EarlinessMA = {}
    LatenessMA = {}
    
    # repeat allocation procedure for all items in the list
    for order in seq['seq']:
        
        item = itemList[order]
#        print 'item', item['orderID']
        
        #================================================
        # Allocation step 1...allocation at current Week
        #================================================
        
        Results = {}
        step = 1
        ind = G.WeekList.index(initialWeek)
        weekList = [initialWeek]
        capacity = deepcopy(G.CurrentCapacityDict)
        qty = item['Qty']
        Allocation = []
        earliness = 0
        lateness = 0
        previousAss = {}
        for ma in G.SPlist[item['sp']]:
            previousAss[ma] = 0

        
        while step <= 3 and qty>0:
            
            if step == 2:                    
                weekList = [G.WeekList[i] for i in range(ind-1, max(-1,ind-G.maxEarliness-1), -1)]
                
            if step == 3:
                weekList = [G.WeekList[i] for i in range(ind+1, min(G.planningHorizon,ind+G.maxLateness+1))]
            
#            print 'weeklist', weekList
            if len(weekList) == 0:
                step+=1
                continue
            
            # check different MAs
            for week in weekList:
                
                # optimise MA allocation               
                spAllocation = Allocation_IP(item, week, previousAss, capacity,G.weightFactor)
#                print 'all', spAllocation
                # implement optimal MA solution
                for ma in spAllocation.keys():
                    if spAllocation[ma]:
                        Results = Allocation2(ma, spAllocation[ma], [week], capacity, G.incompleteBatches, earliness, lateness, Allocation, initialWeek)
                        assert (Results['remainingUnits'] == 0)
                        
                        # update variables
                        capacity = deepcopy(Results['remainingCap'])
                        qty -= spAllocation[ma]
                        Allocation = deepcopy(Results['Allocation'])
                        earliness = deepcopy(Results['earliness'])
                        lateness = deepcopy(Results['lateness'])
                        if ma not in EarlinessMA:
                            EarlinessMA[ma] = 0
                            LatenessMA[ma] = 0
                        EarlinessMA[ma] += max([0, initialWeek - week])*spAllocation[ma]
                        LatenessMA[ma] += max([0, week - initialWeek])*spAllocation[ma]
                            
                        previousAss[ma] += spAllocation[ma]
                        
                if qty <= 0:
                    break
            
            step += 1
        
        # confirm results 
        if qty <= 0:
            G.CurrentCapacityDict = Results['remainingCap']
            G.incompleteBatches = Results['remUnits']
#            print initialWeek, G.Earliness
            for maT in EarlinessMA:
                G.Earliness[initialWeek][maT]['qty'].append(item['Qty'])
                G.Earliness[initialWeek][maT]['earliness'].append(EarlinessMA[maT]/item['Qty'])
                G.Lateness[initialWeek][maT]['qty'].append(item['Qty'])
                G.Lateness[initialWeek][maT]['lateness'].append(LatenessMA[maT]/item['Qty'])
            G.orders[item['orderID']]['Allocation'] = Results['Allocation']
            G.orders[item['orderID']]['Excess'] = False
            chosenMA = []
            for allMA in Results['Allocation']:
                if allMA['ma'] not in chosenMA:
                    chosenMA.append(allMA['ma'])
            G.orders[item['orderID']]['chosenMA'] = chosenMA    
            
                           
    
            for allRep in Results['Allocation']:
                G.globalMAAllocation[allRep['ma']][allRep['week']][itemType][item['priority']] += allRep['units']
            
        
            if itemType == 'forecast':
                G.forecastResults.append((item['ppos'], item['sp'], item['MAlist'], item['Week'], item['Qty'], item['priority'], chosenMA, Results['lateness'], Results['earliness']/item['Qty'], Results['Allocation']))
                
        
        else:
            G.Excess[item['sp']][initialWeek] += item['Qty']
            G.orders[item['orderID']]['Allocation'] = []
            G.orders[item['orderID']]['Excess'] = True
            G.orders[item['orderID']]['chosenMA'] = None
            if itemType == 'forecast':
                G.forecastResults.append((item['ppos'], item['sp'], item['MAlist'], item['Week'], item['Qty'], item['priority'], None, 'NaN', 'NaN', 'None'))
Example #5
0
def AllocationRoutine_Forecast(initialWeek, itemList, itemType, seq):

    EarlinessMA = {}
    LatenessMA = {}

    # repeat allocation procedure for all items in the list
    for order in seq['seq']:

        item = itemList[order]
        #        print 'item', item['orderID']

        #================================================
        # Allocation step 1...allocation at current Week
        #================================================

        Results = {}
        step = 1
        ind = G.WeekList.index(initialWeek)
        weekList = [initialWeek]
        capacity = deepcopy(G.CurrentCapacityDict)
        qty = item['Qty']
        Allocation = []
        earliness = 0
        lateness = 0
        previousAss = {}
        for ma in G.SPlist[item['sp']]:
            previousAss[ma] = 0

        while step <= 3 and qty > 0:

            if step == 2:
                weekList = [
                    G.WeekList[i]
                    for i in range(ind - 1, max(-1, ind - G.maxEarliness -
                                                1), -1)
                ]

            if step == 3:
                weekList = [
                    G.WeekList[i] for i in range(
                        ind + 1, min(G.planningHorizon, ind + G.maxLateness +
                                     1))
                ]

#            print 'weeklist', weekList
            if len(weekList) == 0:
                step += 1
                continue

            # check different MAs
            for week in weekList:

                # optimise MA allocation
                spAllocation = Allocation_IP(item, week, previousAss, capacity,
                                             G.weightFactor)
                #                print 'all', spAllocation
                # implement optimal MA solution
                for ma in spAllocation.keys():
                    if spAllocation[ma]:
                        Results = Allocation2(ma, spAllocation[ma], [week],
                                              capacity, G.incompleteBatches,
                                              earliness, lateness, Allocation,
                                              initialWeek)
                        assert (Results['remainingUnits'] == 0)

                        # update variables
                        capacity = deepcopy(Results['remainingCap'])
                        qty -= spAllocation[ma]
                        Allocation = deepcopy(Results['Allocation'])
                        earliness = deepcopy(Results['earliness'])
                        lateness = deepcopy(Results['lateness'])
                        if ma not in EarlinessMA:
                            EarlinessMA[ma] = 0
                            LatenessMA[ma] = 0
                        EarlinessMA[ma] += max([0, initialWeek - week
                                                ]) * spAllocation[ma]
                        LatenessMA[ma] += max([0, week - initialWeek
                                               ]) * spAllocation[ma]

                        previousAss[ma] += spAllocation[ma]

                if qty <= 0:
                    break

            step += 1

        # confirm results
        if qty <= 0:
            G.CurrentCapacityDict = Results['remainingCap']
            G.incompleteBatches = Results['remUnits']
            #            print initialWeek, G.Earliness
            for maT in EarlinessMA:
                G.Earliness[initialWeek][maT]['qty'].append(item['Qty'])
                G.Earliness[initialWeek][maT]['earliness'].append(
                    EarlinessMA[maT] / item['Qty'])
                G.Lateness[initialWeek][maT]['qty'].append(item['Qty'])
                G.Lateness[initialWeek][maT]['lateness'].append(
                    LatenessMA[maT] / item['Qty'])
            G.orders[item['orderID']]['Allocation'] = Results['Allocation']
            G.orders[item['orderID']]['Excess'] = False
            chosenMA = []
            for allMA in Results['Allocation']:
                if allMA['ma'] not in chosenMA:
                    chosenMA.append(allMA['ma'])
            G.orders[item['orderID']]['chosenMA'] = chosenMA

            for allRep in Results['Allocation']:
                G.globalMAAllocation[allRep['ma']][allRep['week']][itemType][
                    item['priority']] += allRep['units']

            if itemType == 'forecast':
                G.forecastResults.append(
                    (item['ppos'], item['sp'], item['MAlist'], item['Week'],
                     item['Qty'], item['priority'], chosenMA,
                     Results['lateness'], Results['earliness'] / item['Qty'],
                     Results['Allocation']))

        else:
            G.Excess[item['sp']][initialWeek] += item['Qty']
            G.orders[item['orderID']]['Allocation'] = []
            G.orders[item['orderID']]['Excess'] = True
            G.orders[item['orderID']]['chosenMA'] = None
            if itemType == 'forecast':
                G.forecastResults.append(
                    (item['ppos'], item['sp'], item['MAlist'], item['Week'],
                     item['Qty'], item['priority'], None, 'NaN', 'NaN',
                     'None'))
def AllocationRoutine_Forecast(initialWeek, itemList, itemType, seq):

    # repeat allocation procedure for all items in the list
    for order in seq['seq']:

        item = itemList[order]
        #        print 'item', item['orderID']

        EarlinessMA = {}
        LatenessMA = {}
        lateForecast = 0
        earlyForecast = 0
        #================================================
        # Allocation step 1...allocation at current Week
        #================================================

        Results = {}
        step = 1
        ind = G.WeekList.index(initialWeek)
        weekList = [initialWeek]
        weekLateness = [0]
        capacity = deepcopy(G.CurrentCapacityDict)
        incompleteBatches = deepcopy(G.incompleteBatches)
        qty = item['Qty']
        Allocation = []
        earliness = 0
        lateness = 0
        previousAss = {}
        for ma in G.SPlist[item['sp']]:
            previousAss[ma] = 0
            previousAss[ma] = incompleteBatches[ma]
            #            print 'ma in', ma, incompleteBatches[ma]
            qty -= incompleteBatches[ma]
            # make use of incomplete batch units available
            incompleteBatches[ma] = 0

#        print 'order qty', qty

        while step <= 3 and qty > 0:

            if step == 2:
                weekList = [
                    G.WeekList[i]
                    for i in range(ind - 1, max(-1, ind - G.maxEarliness -
                                                1), -1)
                ]
                weekLateness = [
                    i - ind
                    for i in range(ind - 1, max(-1, ind - G.maxEarliness -
                                                1), -1)
                ]
                assert (len(weekList) == len(weekLateness))

            if step == 3:
                weekList = [
                    G.WeekList[i] for i in range(
                        ind + 1, min(G.planningHorizon, ind + G.maxLateness +
                                     1))
                ]
                weekLateness = [
                    i - ind for i in range(
                        ind + 1, min(G.planningHorizon, ind + G.maxLateness +
                                     1))
                ]

#            print 'weeklist', weekList
            if len(weekList) == 0:
                step += 1
                continue

            # check different MAs
            for weekIndex in range(len(weekList)):

                week = weekList[weekIndex]

                # optimise MA allocation
                spAllocation, probStatus = Allocation_IP(
                    item, week, previousAss, capacity, G.weightFactor)
                #                print 'all', spAllocation
                # implement optimal MA solution
                for ma in spAllocation.keys():
                    if spAllocation[ma]:
                        Results = Allocation2(ma, spAllocation[ma], [week],
                                              capacity, incompleteBatches,
                                              earliness, lateness, Allocation,
                                              initialWeek)
                        assert (Results['remainingUnits'] == 0)

                        # update variables
                        capacity = deepcopy(Results['remainingCap'])
                        qty -= spAllocation[ma]
                        Allocation = deepcopy(Results['Allocation'])
                        earliness = deepcopy(Results['earliness'])
                        lateness = deepcopy(Results['lateness'])
                        if ma not in EarlinessMA:
                            EarlinessMA[ma] = 0
                            LatenessMA[ma] = 0

                        EarlinessMA[ma] += max(
                            [0, weekLateness[weekIndex] *
                             (-1)]) * spAllocation[ma]
                        LatenessMA[ma] += max([
                            0, weekLateness[weekIndex]
                        ]) * spAllocation[ma]  #week - initialWeek

                        previousAss[ma] += spAllocation[ma]

                if qty <= 0:
                    break

            step += 1

        # confirm results
        if qty <= 0:

            minAss = item['Qty'] * 10
            maIB = None
            G.CurrentCapacityDict = Results['remainingCap']
            G.incompleteBatches = Results['remUnits']
            #            print initialWeek, G.Earliness
            for maT in EarlinessMA:
                #                assert(Results['remUnits'][maT]==0)
                G.Earliness[initialWeek][maT]['qty'].append(item['Qty'])
                G.Earliness[initialWeek][maT]['earliness'].append(
                    EarlinessMA[maT] / item['Qty'])
                G.Lateness[initialWeek][maT]['qty'].append(item['Qty'])
                G.Lateness[initialWeek][maT]['lateness'].append(
                    LatenessMA[maT] / item['Qty'])
                lateForecast += LatenessMA[maT] / item['Qty']
                earlyForecast += EarlinessMA[maT] / item['Qty']
                if previousAss[maT] <= minAss:
                    minAss = previousAss[maT]
                    maIB = maT

#            print 'ma ib', maIB, qty, cQty, item['Qty'], previousAss, EarlinessMA , [G.incompleteBatches[ma] for ma in G.SPlist[item['sp']]]
            if maIB != None:
                G.incompleteBatches[maIB] -= qty
#            print [G.incompleteBatches[ma] for ma in G.SPlist[item['sp']]]

            G.orders[item['orderID']]['Allocation'] = Results['Allocation']
            G.orders[item['orderID']]['Excess'] = False
            chosenMA = []
            for allMA in Results['Allocation']:
                if allMA['ma'] not in chosenMA:
                    chosenMA.append(allMA['ma'])
            G.orders[item['orderID']]['chosenMA'] = chosenMA

            if lateForecast:
                G.LateMeasures['noLateOrders'] += 1
                G.LateMeasures['lateness'].append(lateForecast)

            if earlyForecast:
                G.LateMeasures['noEarlyOrders'] += 1
                G.LateMeasures['earliness'].append(earlyForecast)

            for allRep in Results['Allocation']:
                G.globalMAAllocation[allRep['ma']][allRep['week']][itemType][
                    item['priority']] += allRep['units']
                G.globalMAAllocationIW[allRep['ma']][initialWeek][itemType][
                    item['priority']] += allRep['units']

            if itemType == 'forecast':
                G.forecastResults.append(
                    (item['ppos'], item['sp'], item['MAlist'], item['Week'],
                     item['Qty'], item['priority'], chosenMA,
                     Results['lateness'] / item['Qty'], Results['earliness'] /
                     item['Qty'], Results['Allocation']))

        else:
            G.Excess[item['sp']][initialWeek] += item['Qty']
            G.orders[item['orderID']]['Allocation'] = []
            G.orders[item['orderID']]['Excess'] = True
            G.orders[item['orderID']]['chosenMA'] = None
            G.globalMAAllocationIW[item['sp']][initialWeek][itemType][
                item['priority']] += item['Qty']
            G.LateMeasures['noExcess'] += 1
            G.LateMeasures['exUnits'] += item['Qty']
            if itemType == 'forecast':
                G.forecastResults.append(
                    (item['ppos'], item['sp'], item['MAlist'], item['Week'],
                     item['Qty'], item['priority'], None, 'NaN', 'NaN',
                     'None'))