// createCoupleGeneticAnalysis provides functions to create a Couple genetic analysis
// Couple analyses provide an analysis of the prospective offspring of a pair of people
// These analyses are performed on one or more genome files.
// Analyses contain 3 categories of results: Monogenic Diseases, Polygenic Diseases and Traits
// Use createPersonGeneticAnalysis.go to create Person analyses

package createCoupleGeneticAnalysis

// Disclaimer: I am a novice in the ways of genetics. This package could be flawed in numerous ways.

// TODO: We want to eventually use neural nets for polygenic disease analysis (see geneticPrediction.go)
// This is only possible once we get access to the necessary training data

import "seekia/resources/trainedPredictionModels"
import "seekia/resources/geneticReferences/locusMetadata"
import "seekia/resources/geneticReferences/monogenicDiseases"
import "seekia/resources/geneticReferences/polygenicDiseases"
import "seekia/resources/geneticReferences/traits"

import "seekia/internal/encoding"
import "seekia/internal/genetics/createPersonGeneticAnalysis"
import "seekia/internal/genetics/geneticAnalysis"
import "seekia/internal/genetics/locusValue"
import "seekia/internal/genetics/prepareRawGenomes"
import "seekia/internal/helpers"

import "errors"
import mathRand "math/rand/v2"
import "maps"
import "reflect"


//Outputs:
//	-bool: Process completed (was not manually stopped mid-way)
//	-string: Couple genetic analysis string (encoded in MessagePack)
//	-error
func CreateCoupleGeneticAnalysis(person1GenomesList []prepareRawGenomes.RawGenomeWithMetadata, person2GenomesList []prepareRawGenomes.RawGenomeWithMetadata, updatePercentageCompleteFunction func(int)error, checkIfProcessIsStopped func()bool)(bool, string, error){

	person1PrepareRawGenomesUpdatePercentageCompleteFunction := func(newPercentage int)error{

		newPercentageCompletion, err := helpers.ScaleIntProportionally(true, newPercentage, 0, 100, 0, 25)
		if (err != nil) { return err }

		err = updatePercentageCompleteFunction(newPercentageCompletion)
		if (err != nil) { return err }

		return nil
	}

	anyUsefulLocationsExist, person1GenomesWithMetadataList, allPerson1RawGenomeIdentifiersList, person1HasMultipleGenomes, person1OnlyExcludeConflictsGenomeIdentifier, person1OnlyIncludeSharedGenomeIdentifier, err := prepareRawGenomes.GetGenomesWithMetadataListFromRawGenomesList(person1GenomesList, person1PrepareRawGenomesUpdatePercentageCompleteFunction)
	if (err != nil) { return false, "", err }
	if (anyUsefulLocationsExist == false){
		// We should have checked for this when genomes were first imported.
		return false, "", errors.New("CreateCoupleGeneticAnalysis called with person1GenomesList that does not contain any useful genomes")
	}
	if (len(person1GenomesList) > 1 && (len(person1GenomesList) != (len(person1GenomesWithMetadataList)-2)) ){
		// If there is more than 1 genome, 2 combined genomes are created
		// We are checking to make sure that none of the input genomes were dropped due to not having any locations
		// We should have checked to make sure each input genome has useful locations when each genome was first imported.
		return false, "", errors.New("CreateCoupleGeneticAnalysis called with person1GenomesList containing at least 1 genome without useful locations.")
	}

	processIsStopped := checkIfProcessIsStopped()
	if (processIsStopped == true){
		return false, "", nil
	}

	person2PrepareRawGenomesUpdatePercentageCompleteFunction := func(newPercentage int)error{

		newPercentageCompletion, err := helpers.ScaleIntProportionally(true, newPercentage, 0, 100, 25, 50)
		if (err != nil){ return err }

		err = updatePercentageCompleteFunction(newPercentageCompletion)
		if (err != nil) { return err }

		return nil
	}

	anyUsefulLocationsExist, person2GenomesWithMetadataList, allPerson2RawGenomeIdentifiersList, person2HasMultipleGenomes, person2OnlyExcludeConflictsGenomeIdentifier, person2OnlyIncludeSharedGenomeIdentifier, err := prepareRawGenomes.GetGenomesWithMetadataListFromRawGenomesList(person2GenomesList, person2PrepareRawGenomesUpdatePercentageCompleteFunction)
	if (err != nil) { return false, "", err }
	if (anyUsefulLocationsExist == false){
		// We should have checked for this when genomes were first imported.
		return false, "", errors.New("CreateCoupleGeneticAnalysis called with person2GenomesList that does not contain any useful genomes")
	}
	if (len(person2GenomesList) > 1 && (len(person2GenomesList) != (len(person2GenomesWithMetadataList)-2)) ){
		// If there is more than 1 genome, 2 combined genomes are created
		// We are checking to make sure that none of the input genomes were dropped due to not having any locations
		// We should have checked to make sure each input genome has useful locations when each genome was first imported.
		return false, "", errors.New("CreateCoupleGeneticAnalysis called with person2GenomesList containing at least 1 genome without useful locations.")
	}

	processIsStopped = checkIfProcessIsStopped()
	if (processIsStopped == true){
		return false, "", nil
	}

	// This map stores each genome's locus values
	// Map Structure: Genome Identifier -> Genome locus values map (rsID -> Locus Value)
	person1GenomesMap := make(map[[16]byte]map[int64]locusValue.LocusValue)

	for _, genomeWithMetadata := range person1GenomesWithMetadataList{

		genomeIdentifier := genomeWithMetadata.GenomeIdentifier
		genomeMap := genomeWithMetadata.GenomeMap

		person1GenomesMap[genomeIdentifier] = genomeMap
	}

	// This map stores each genome's locus values
	// Map Structure: Genome Identifier -> Genome locus values map (rsID -> Locus Value)
	person2GenomesMap := make(map[[16]byte]map[int64]locusValue.LocusValue)

	for _, genomeWithMetadata := range person2GenomesWithMetadataList{

		genomeIdentifier := genomeWithMetadata.GenomeIdentifier
		genomeMap := genomeWithMetadata.GenomeMap

		person2GenomesMap[genomeIdentifier] = genomeMap
	}

	// The analysis will analyze either 1 or 2 genome pairs
	// The gui will display the results from each pair
	//Outputs:
	//	-[16]byte: Pair 1 Person 1 Genome Identifier
	//	-[16]byte: Pair 1 Person 2 Genome Identifier
	//	-bool: Second pair exists
	//	-[16]byte: Pair 2 Person 1 Genome Identifier
	//	-[16]byte: Pair 2 Person 2 Genome Identifier
	//	-error
	getGenomePairsToAnalyze := func()([16]byte, [16]byte, bool, [16]byte, [16]byte, error){

		if (person1HasMultipleGenomes == true && person2HasMultipleGenomes == true){

			return person1OnlyExcludeConflictsGenomeIdentifier, person2OnlyExcludeConflictsGenomeIdentifier, true, person1OnlyIncludeSharedGenomeIdentifier, person2OnlyIncludeSharedGenomeIdentifier, nil
		}
		if (person1HasMultipleGenomes == true && person2HasMultipleGenomes == false){

			person2GenomeIdentifier := allPerson2RawGenomeIdentifiersList[0]

			return person1OnlyExcludeConflictsGenomeIdentifier, person2GenomeIdentifier, true, person1OnlyIncludeSharedGenomeIdentifier, person2GenomeIdentifier, nil
		}
		if (person1HasMultipleGenomes == false && person2HasMultipleGenomes == true){

			person1GenomeIdentifier := allPerson1RawGenomeIdentifiersList[0]

			return person1GenomeIdentifier, person2OnlyExcludeConflictsGenomeIdentifier, true, person1GenomeIdentifier, person2OnlyIncludeSharedGenomeIdentifier, nil
		}

		//person1HasMultipleGenomes == false && person2HasMultipleGenomes == false

		person1GenomeIdentifier := allPerson1RawGenomeIdentifiersList[0]
		person2GenomeIdentifier := allPerson2RawGenomeIdentifiersList[0]

		return person1GenomeIdentifier, person2GenomeIdentifier, false, [16]byte{}, [16]byte{}, nil
	}

	pair1Person1GenomeIdentifier, pair1Person2GenomeIdentifier, genomePair2Exists, pair2Person1GenomeIdentifier, pair2Person2GenomeIdentifier, err := getGenomePairsToAnalyze()
	if (err != nil){ return false, "", err }

	newCoupleAnalysis := geneticAnalysis.CoupleAnalysis{
		AnalysisVersion: 1,
		Pair1Person1GenomeIdentifier: pair1Person1GenomeIdentifier,
		Pair1Person2GenomeIdentifier: pair1Person2GenomeIdentifier,
		SecondPairExists: genomePair2Exists,
		Person1HasMultipleGenomes: person1HasMultipleGenomes,
		Person2HasMultipleGenomes: person2HasMultipleGenomes,
	}

	if (genomePair2Exists == true){

		// At least 1 of the people have multiple genomes

		newCoupleAnalysis.Pair2Person1GenomeIdentifier = pair2Person1GenomeIdentifier
		newCoupleAnalysis.Pair2Person2GenomeIdentifier = pair2Person2GenomeIdentifier

		if (person1HasMultipleGenomes == true){
			newCoupleAnalysis.Person1OnlyExcludeConflictsGenomeIdentifier = person1OnlyExcludeConflictsGenomeIdentifier
			newCoupleAnalysis.Person1OnlyIncludeSharedGenomeIdentifier = person1OnlyIncludeSharedGenomeIdentifier
		}

		if (person2HasMultipleGenomes == true){
			newCoupleAnalysis.Person2OnlyExcludeConflictsGenomeIdentifier = person2OnlyExcludeConflictsGenomeIdentifier
			newCoupleAnalysis.Person2OnlyIncludeSharedGenomeIdentifier = person2OnlyIncludeSharedGenomeIdentifier
		}
	}

	// We compute and add monogenic disease probabilities

	monogenicDiseasesList, err := monogenicDiseases.GetMonogenicDiseaseObjectsList()
	if (err != nil) { return false, "", err }

	// Map Structure: Disease Name -> OffspringMonogenicDiseaseInfo
	offspringMonogenicDiseasesMap := make(map[string]geneticAnalysis.OffspringMonogenicDiseaseInfo)

	for _, diseaseObject := range monogenicDiseasesList{

		diseaseName := diseaseObject.DiseaseName
		variantsList := diseaseObject.VariantsList
		diseaseIsDominantOrRecessive := diseaseObject.DominantOrRecessive

		person1DiseaseAnalysisObject, err := createPersonGeneticAnalysis.GetPersonMonogenicDiseaseAnalysis(person1GenomesWithMetadataList, diseaseObject)
		if (err != nil) { return false, "", err }

		person2DiseaseAnalysisObject, err := createPersonGeneticAnalysis.GetPersonMonogenicDiseaseAnalysis(person2GenomesWithMetadataList, diseaseObject)
		if (err != nil) { return false, "", err }

		// This map stores the quantity of variants tested in each person's genome
		// Map Structure: Genome Identifier -> Number of variants tested
		quantityOfVariantsTestedMap := make(map[[16]byte]int)

		// This map stores the offspring disease probabilities for each genome pair.
		// A genome pair is a concatenation of two genome identifiers
		// If a map entry doesn't exist, the probabilities are unknown for that genome pair
		// Map Structure: Genome Pair Identifier -> OffspringMonogenicDiseaseProbabilities
		offspringMonogenicDiseaseInfoMap := make(map[[32]byte]geneticAnalysis.OffspringGenomePairMonogenicDiseaseInfo)

		// This will calculate the probability of monogenic disease for the offspring from the two specified genomes
		// It also calculates the probabilities for each monogenic disease variant for the offspring
		// It then adds the genome pair disease information to the offspringMonogenicDiseaseInfoMap and quantityOfVariantsTestedMap
		addGenomePairInfoToDiseaseMaps := func(person1GenomeIdentifier [16]byte, person2GenomeIdentifier [16]byte)error{

			//Outputs:
			//	-bool: Probability is known
			//	-int: Probability of passing a disease variant (value between 0 and 100)
			//	-int: Number of variants tested
			getPersonWillPassDiseaseVariantProbability := func(personDiseaseAnalysisObject geneticAnalysis.PersonMonogenicDiseaseInfo, genomeIdentifier [16]byte)(bool, int, int){

				personDiseaseInfoMap := personDiseaseAnalysisObject.MonogenicDiseaseInfoMap

				personGenomeDiseaseInfoObject, exists := personDiseaseInfoMap[genomeIdentifier]
				if (exists == false){
					return false, 0, 0
				}

				personGenomeProbabilityOfPassingADiseaseVariant := personGenomeDiseaseInfoObject.ProbabilityOfPassingADiseaseVariant

				personGenomeQuantityOfVariantsTested := personGenomeDiseaseInfoObject.QuantityOfVariantsTested

				return true, personGenomeProbabilityOfPassingADiseaseVariant, personGenomeQuantityOfVariantsTested
			}

			person1ProbabilityIsKnown, person1WillPassVariantProbability, person1NumberOfVariantsTested := getPersonWillPassDiseaseVariantProbability(person1DiseaseAnalysisObject, person1GenomeIdentifier)

			person2ProbabilityIsKnown, person2WillPassVariantProbability, person2NumberOfVariantsTested := getPersonWillPassDiseaseVariantProbability(person2DiseaseAnalysisObject, person2GenomeIdentifier)

			offspringHasDiseaseProbabilityIsKnown, percentageProbabilityOffspringHasDisease, offspringHasAVariantProbabilityIsKnown, percentageProbabilityOffspringHasVariant, err := GetOffspringMonogenicDiseaseProbabilities(diseaseIsDominantOrRecessive, person1ProbabilityIsKnown, person1WillPassVariantProbability, person2ProbabilityIsKnown, person2WillPassVariantProbability)
			if (err != nil) { return err }

			// Now we calculate the probabilities for individual variants

			offspringVariantsInfoMap := make(map[[3]byte]geneticAnalysis.OffspringMonogenicDiseaseVariantInfo)

			for _, variantObject := range variantsList{

				variantIdentifierHex := variantObject.VariantIdentifier

				variantIdentifier, err := encoding.DecodeHexStringTo3ByteArray(variantIdentifierHex)
				if (err != nil) { return err }

				//Outputs:
				//	-bool: Probability is known
				//	-float64: Probability that person will pass variant to offspring (between 0 and 1)
				//	-error
				getPersonWillPassVariantProbability := func(personDiseaseAnalysisObject geneticAnalysis.PersonMonogenicDiseaseInfo, personGenomeIdentifier [16]byte)(bool, float64, error){

					personDiseaseInfoMap := personDiseaseAnalysisObject.MonogenicDiseaseInfoMap

					personGenomeDiseaseInfoObject, exists := personDiseaseInfoMap[personGenomeIdentifier]
					if (exists == false){
						return false, 0, nil
					}

					personGenomeDiseaseVariantsMap := personGenomeDiseaseInfoObject.VariantsInfoMap

					personVariantInfoObject, exists := personGenomeDiseaseVariantsMap[variantIdentifier]
					if (exists == false){
						// The genome does not have information about this variant

						return false, 0, nil
					}

					base1HasVariant := personVariantInfoObject.Base1HasVariant
					base2HasVariant := personVariantInfoObject.Base2HasVariant

					if (base1HasVariant == true && base2HasVariant == true){
						return true, 1, nil
					}
					if (base1HasVariant == true && base2HasVariant == false){
						return true, 0.5, nil
					}
					if (base1HasVariant == false && base2HasVariant == true){
						return true, 0.5, nil
					}

					// Neither base has a variant

					return true, 0, nil
				}

				person1VariantProbabilityIsKnown, person1WillPassVariantProbability, err := getPersonWillPassVariantProbability(person1DiseaseAnalysisObject, person1GenomeIdentifier)
				if (err != nil) { return err }

				person2VariantProbabilityIsKnown, person2WillPassVariantProbability, err := getPersonWillPassVariantProbability(person2DiseaseAnalysisObject, person2GenomeIdentifier)
				if (err != nil) { return err }

				if (person1VariantProbabilityIsKnown == false && person2VariantProbabilityIsKnown == false){
					continue
				}

				// Outputs:
				//	-float64: Best Case Person 1 will pass variant probability (0-1)
				//	-float64: Worst Case Person 1 will pass variant probability (0-1)
				//	-float64: Best Case Person 2 will pass variant probability (0-1)
				//	-float64: Worst Case Person 2 will pass variant probability (0-1)
				getBestAndWorstCaseProbabilities := func()(float64, float64, float64, float64){

					if (person1VariantProbabilityIsKnown == true && person2VariantProbabilityIsKnown == false){

						return person1WillPassVariantProbability, person1WillPassVariantProbability, float64(0), float64(1)
					}
					if (person1VariantProbabilityIsKnown == false && person2VariantProbabilityIsKnown == true){
						return float64(0), float64(1), person2WillPassVariantProbability, person2WillPassVariantProbability
					}

					// Both people's probabilities are known
					return person1WillPassVariantProbability, person1WillPassVariantProbability, person2WillPassVariantProbability, person2WillPassVariantProbability
				}

				bestCasePerson1WillPassVariantProbability, worstCasePerson1WillPassVariantProbability, bestCasePerson2WillPassVariantProbability, worstCasePerson2WillPassVariantProbability := getBestAndWorstCaseProbabilities()

				//Outputs:
				//	-int: Percentage Probability of 0 mutations
				//	-int: Percentage Probability of 1 mutation
				//	-int: Percentage Probability of 2 mutations
				//	-error
				getOffspringVariantMutationProbabilities := func(person1WillPassVariantProbability float64, person2WillPassVariantProbability float64)(int, int, int, error){

					// This is the probability that neither person will pass the variant
					// P = P(!A) * P(!B)
					probabilityOf0Mutations := (1 - person1WillPassVariantProbability) * (1 - person2WillPassVariantProbability)

					// This is the probability that either person will pass the variant, but not both
					// P(A XOR B) = P(A) + P(B) - (2 * P(A and B))
					probabilityOf1Mutation := person1WillPassVariantProbability + person2WillPassVariantProbability - (2 * person1WillPassVariantProbability * person2WillPassVariantProbability)

					// This is the probability that both people will pass the variant
					// P(A and B) = P(A) * P(B)
					probabilityOf2Mutations := person1WillPassVariantProbability * person2WillPassVariantProbability

					percentageProbabilityOf0Mutations, err := helpers.FloorFloat64ToInt(probabilityOf0Mutations * 100)
					if (err != nil) { return 0, 0, 0, err }
					percentageProbabilityOf1Mutation, err := helpers.FloorFloat64ToInt(probabilityOf1Mutation * 100)
					if (err != nil) { return 0, 0, 0, err }
					percentageProbabilityOf2Mutations, err := helpers.FloorFloat64ToInt(probabilityOf2Mutations * 100)
					if (err != nil) { return 0, 0, 0, err }

					return percentageProbabilityOf0Mutations, percentageProbabilityOf1Mutation, percentageProbabilityOf2Mutations, nil
				}

				bestCase0MutationsProbability, bestCase1MutationProbability, bestCase2MutationsProbability, err := getOffspringVariantMutationProbabilities(bestCasePerson1WillPassVariantProbability, bestCasePerson2WillPassVariantProbability)
				if (err != nil) { return err }

				worstCase0MutationsProbability, worstCase1MutationProbability, worstCase2MutationsProbability, err := getOffspringVariantMutationProbabilities(worstCasePerson1WillPassVariantProbability, worstCasePerson2WillPassVariantProbability)
				if (err != nil) { return err }

				// We have to figure out which 1-mutation-probability is lower
				// The best case probabilities can actually result in a higher probability for 1 mutation
				//		than the worst case probabilities
				// This is because in a worst-case-scenario, the probability of 1 mutation might be 0 because
				//		the probability of 2 mutations is 100
				// This is not needed for the 0 and 2 mutation probabilities because TODO

				lowerBound1MutationProbability := min(bestCase1MutationProbability, worstCase1MutationProbability)
				upperBound1MutationProbability := max(bestCase1MutationProbability, worstCase1MutationProbability)

				newOffspringMonogenicDiseaseVariantInfoObject := geneticAnalysis.OffspringMonogenicDiseaseVariantInfo{

					ProbabilityOf0MutationsLowerBound: worstCase0MutationsProbability,
					ProbabilityOf0MutationsUpperBound: bestCase0MutationsProbability,

					ProbabilityOf1MutationLowerBound: lowerBound1MutationProbability,
					ProbabilityOf1MutationUpperBound: upperBound1MutationProbability,

					ProbabilityOf2MutationsLowerBound: bestCase2MutationsProbability,
					ProbabilityOf2MutationsUpperBound: worstCase2MutationsProbability,
				}

				offspringVariantsInfoMap[variantIdentifier] = newOffspringMonogenicDiseaseVariantInfoObject
			}

			if (person1ProbabilityIsKnown == false && person2ProbabilityIsKnown == false && len(offspringVariantsInfoMap) == 0){
				// We have no information about this genome pair's disease risk
				// We don't add this genome pair's disease info to the diseaseInfoMap
				return nil
			}

			quantityOfVariantsTestedMap[person1GenomeIdentifier] = person1NumberOfVariantsTested
			quantityOfVariantsTestedMap[person2GenomeIdentifier] = person2NumberOfVariantsTested

			newOffspringGenomePairMonogenicDiseaseInfoObject := geneticAnalysis.OffspringGenomePairMonogenicDiseaseInfo{
				ProbabilityOffspringHasDiseaseIsKnown: offspringHasDiseaseProbabilityIsKnown,
				ProbabilityOffspringHasDisease: percentageProbabilityOffspringHasDisease,
				ProbabilityOffspringHasVariantIsKnown: offspringHasAVariantProbabilityIsKnown,
				ProbabilityOffspringHasVariant: percentageProbabilityOffspringHasVariant,
				VariantsInfoMap: offspringVariantsInfoMap,
			}

			genomePairIdentifier := helpers.JoinTwo16ByteArrays(person1GenomeIdentifier, person2GenomeIdentifier)

			offspringMonogenicDiseaseInfoMap[genomePairIdentifier] = newOffspringGenomePairMonogenicDiseaseInfoObject

			return nil
		}

		err = addGenomePairInfoToDiseaseMaps(pair1Person1GenomeIdentifier, pair1Person2GenomeIdentifier)
		if (err != nil) { return false, "", err }

		if (genomePair2Exists == true){

			err := addGenomePairInfoToDiseaseMaps(pair2Person1GenomeIdentifier, pair2Person2GenomeIdentifier)
			if (err != nil) { return false, "", err }
		}

		newOffspringMonogenicDiseaseInfoObject := geneticAnalysis.OffspringMonogenicDiseaseInfo{
			QuantityOfVariantsTestedMap: quantityOfVariantsTestedMap,
			MonogenicDiseaseInfoMap: offspringMonogenicDiseaseInfoMap,
		}

		// Now we check for conflicts in monogenic disease results

		// For couples, a conflict is when either genome pair has differing results for any disease probability/variant probability
		// This is different from a person having conflicts within their different genomes
		// Each genome pair only uses 1 genome from each person.

		if (len(offspringMonogenicDiseaseInfoMap) >= 2){

			// Conflicts are only possible if two genome pairs with information about this disease exist

			checkIfConflictExists := func()(bool, error){

				probabilityOffspringHasDiseaseIsKnown := false
				probabilityOffspringHasDisease := 0
				probabilityOffspringHasVariantIsKnown := false
				probabilityOffspringHasVariant := 0

				offspringVariantsInfoMap := make(map[[3]byte]geneticAnalysis.OffspringMonogenicDiseaseVariantInfo)

				firstItemReached := false

				for _, offspringMonogenicDiseaseInfoObject := range offspringMonogenicDiseaseInfoMap{

					currentProbabilityOffspringHasDiseaseIsKnown := offspringMonogenicDiseaseInfoObject.ProbabilityOffspringHasDiseaseIsKnown
					currentProbabilityOffspringHasDisease := offspringMonogenicDiseaseInfoObject.ProbabilityOffspringHasDisease
					currentProbabilityOffspringHasVariantIsKnown := offspringMonogenicDiseaseInfoObject.ProbabilityOffspringHasVariantIsKnown
					currentProbabilityOffspringHasVariant := offspringMonogenicDiseaseInfoObject.ProbabilityOffspringHasVariant

					currentOffspringVariantsInfoMap := offspringMonogenicDiseaseInfoObject.VariantsInfoMap

					if (firstItemReached == false){
						probabilityOffspringHasDiseaseIsKnown = currentProbabilityOffspringHasDiseaseIsKnown
						probabilityOffspringHasDisease = currentProbabilityOffspringHasDisease
						probabilityOffspringHasVariantIsKnown = currentProbabilityOffspringHasVariantIsKnown
						probabilityOffspringHasVariant = currentProbabilityOffspringHasVariant

						offspringVariantsInfoMap = currentOffspringVariantsInfoMap

						firstItemReached = true
						continue
					}
					if (currentProbabilityOffspringHasDiseaseIsKnown != probabilityOffspringHasDiseaseIsKnown){
						return true, nil
					}
					if (currentProbabilityOffspringHasDisease != probabilityOffspringHasDisease){
						return true, nil
					}
					if (currentProbabilityOffspringHasVariantIsKnown != probabilityOffspringHasVariantIsKnown){
						return true, nil
					}
					if (currentProbabilityOffspringHasVariant != probabilityOffspringHasVariant){
						return true, nil
					}
					areEqual := maps.Equal(currentOffspringVariantsInfoMap, offspringVariantsInfoMap)
					if (areEqual == false){
						return true, nil
					}
				}

				return false, nil
			}

			conflictExists, err := checkIfConflictExists()
			if (err != nil) { return false, "", err }

			newOffspringMonogenicDiseaseInfoObject.ConflictExists = conflictExists
		}

		offspringMonogenicDiseasesMap[diseaseName] = newOffspringMonogenicDiseaseInfoObject
	}

	newCoupleAnalysis.MonogenicDiseasesMap = offspringMonogenicDiseasesMap

	// Step 2: Polygenic Diseases

	polygenicDiseaseObjectsList, err := polygenicDiseases.GetPolygenicDiseaseObjectsList()
	if (err != nil){ return false, "", err }

	// Map Structure: Disease Name -> OffspringPolygenicDiseaseInfo
	offspringPolygenicDiseasesMap := make(map[string]geneticAnalysis.OffspringPolygenicDiseaseInfo)

	for _, diseaseObject := range polygenicDiseaseObjectsList{

		diseaseName := diseaseObject.DiseaseName

		// This map stores the polygenic disease info for each genome pair
		// Map Structure: Genome Pair Identifier -> OffspringGenomePairPolygenicDiseaseInfo
		offspringPolygenicDiseaseInfoMap := make(map[[32]byte]geneticAnalysis.OffspringGenomePairPolygenicDiseaseInfo)

		// This will calculate the disease info for the offspring from the two specified genomes to the diseases map
		// It then adds the pair entry to the offspringPolygenicDiseaseInfoMap
		addGenomePairDiseaseInfoToDiseaseMap := func(person1GenomeIdentifier [16]byte, person2GenomeIdentifier [16]byte)error{

			person1LocusValuesMap, exists := person1GenomesMap[person1GenomeIdentifier]
			if (exists == false){
				return errors.New("addGenomePairDiseaseInfoToDiseaseMap called with unknown person1GenomeIdentifier.")
			}

			person2LocusValuesMap, exists := person2GenomesMap[person2GenomeIdentifier]
			if (exists == false){
				return errors.New("addGenomePairDiseaseInfoToDiseaseMap called with unknown person2GenomeIdentifier.")
			}

			neuralNetworkExists, anyOffspringLocusKnown, offspringAverageRiskScore, accuracyRangesMap, predictedRiskScoresList, quantityOfLociKnown, quantityOfParentalPhasedLoci, err := GetOffspringPolygenicDiseaseAnalysis(diseaseObject, person1LocusValuesMap, person2LocusValuesMap)
			if (err != nil) { return err }
			if (neuralNetworkExists == false){
				// We cannot analyze this disease
				return nil
			}
			if (anyOffspringLocusKnown == false){
				// We have no information about this genome pair's disease risk
				// We don't add this genome pair's disease info to the diseaseInfoMap
				return nil
			}

			newOffspringGenomePairPolygenicDiseaseInfo := geneticAnalysis.OffspringGenomePairPolygenicDiseaseInfo{

				OffspringAverageRiskScore: offspringAverageRiskScore,
				PredictionConfidenceRangesMap: accuracyRangesMap,
				QuantityOfLociKnown: quantityOfLociKnown,
				QuantityOfParentalPhasedLoci: quantityOfParentalPhasedLoci,
				SampleOffspringRiskScoresList: predictedRiskScoresList,
			}

			genomePairIdentifier := helpers.JoinTwo16ByteArrays(person1GenomeIdentifier, person2GenomeIdentifier)

			offspringPolygenicDiseaseInfoMap[genomePairIdentifier] = newOffspringGenomePairPolygenicDiseaseInfo

			return nil
		}

		err = addGenomePairDiseaseInfoToDiseaseMap(pair1Person1GenomeIdentifier, pair1Person2GenomeIdentifier)
		if (err != nil) { return false, "", err }

		if (genomePair2Exists == true){

			err := addGenomePairDiseaseInfoToDiseaseMap(pair2Person1GenomeIdentifier, pair2Person2GenomeIdentifier)
			if (err != nil) { return false, "", err }
		}

		if (len(offspringPolygenicDiseaseInfoMap) == 0){
			// No disease analysis was performed
			continue
		}

		newOffspringPolygenicDiseaseInfoObject := geneticAnalysis.OffspringPolygenicDiseaseInfo{
			PolygenicDiseaseInfoMap: offspringPolygenicDiseaseInfoMap,
		}

		if (len(offspringPolygenicDiseaseInfoMap) >= 2){

			// We check for conflicts
			// Conflicts are only possible if two genome pairs with disease info for this disease exist

			checkIfConflictExists := func()(bool, error){

				currentGenomePairPolygenicDiseaseInfo := geneticAnalysis.OffspringGenomePairPolygenicDiseaseInfo{}

				firstItemReached := false

				for _, genomePairDiseaseInfoObject := range offspringPolygenicDiseaseInfoMap{

					if (firstItemReached == false){
						currentGenomePairPolygenicDiseaseInfo = genomePairDiseaseInfoObject
						firstItemReached = true
						continue
					}

					areEqual := reflect.DeepEqual(genomePairDiseaseInfoObject, currentGenomePairPolygenicDiseaseInfo)
					if (areEqual == false){
						return true, nil
					}
				}

				return false, nil
			}

			conflictExists, err := checkIfConflictExists()
			if (err != nil) { return false, "", err }

			newOffspringPolygenicDiseaseInfoObject.ConflictExists = conflictExists
		}

		offspringPolygenicDiseasesMap[diseaseName] = newOffspringPolygenicDiseaseInfoObject
	}

	newCoupleAnalysis.PolygenicDiseasesMap = offspringPolygenicDiseasesMap

	// Step 3: Traits

	traitObjectsList, err := traits.GetTraitObjectsList()
	if (err != nil) { return false, "", err }

	// Map Structure: Trait Name -> Trait Info Object
	offspringDiscreteTraitsMap := make(map[string]geneticAnalysis.OffspringDiscreteTraitInfo)

	// Map Structure: Trait Name -> Trait Info Object
	offspringNumericTraitsMap := make(map[string]geneticAnalysis.OffspringNumericTraitInfo)

	for _, traitObject := range traitObjectsList{

		traitName := traitObject.TraitName
		traitIsDiscreteOrNumeric := traitObject.DiscreteOrNumeric

		if (traitIsDiscreteOrNumeric == "Discrete"){

			// This map stores the trait info for each genome pair
			// Map Structure: Genome Pair Identifier -> OffspringGenomePairDiscreteTraitInfo
			offspringTraitInfoMap := make(map[[32]byte]geneticAnalysis.OffspringGenomePairDiscreteTraitInfo)

			// This will add the offspring trait information for the provided genome pair to the offspringTraitInfoMap
			addGenomePairTraitInfoToOffspringMap := func(person1GenomeIdentifier [16]byte, person2GenomeIdentifier [16]byte)error{

				person1LocusValuesMap, exists := person1GenomesMap[person1GenomeIdentifier]
				if (exists == false){
					return errors.New("addGenomePairTraitInfoToOffspringMap called with unknown person1GenomeIdentifier.")
				}

				person2LocusValuesMap, exists := person2GenomesMap[person2GenomeIdentifier]
				if (exists == false){
					return errors.New("addGenomePairTraitInfoToOffspringMap called with unknown person2GenomeIdentifier.")
				}

				newOffspringGenomePairTraitInfo := geneticAnalysis.OffspringGenomePairDiscreteTraitInfo{}

				neuralNetworkExists, neuralNetworkAnalysisExists, outcomeProbabilitiesMap, averagePredictionConfidence, quantityOfLociTested, quantityOfParentalPhasedLoci, err := GetOffspringDiscreteTraitAnalysis_NeuralNetwork(traitObject, person1LocusValuesMap, person2LocusValuesMap)
				if (err != nil) { return err }
				if (neuralNetworkExists == true){

					newOffspringGenomePairTraitInfo.NeuralNetworkExists = true

					if (neuralNetworkAnalysisExists == true){
						newOffspringGenomePairTraitInfo.NeuralNetworkAnalysisExists = true

						newOffspringGenomePairTraitInfo_NeuralNetwork := geneticAnalysis.OffspringGenomePairDiscreteTraitInfo_NeuralNetwork{

							OffspringOutcomeProbabilitiesMap: outcomeProbabilitiesMap,
							AverageConfidence: averagePredictionConfidence,
							QuantityOfLociKnown: quantityOfLociTested,
							QuantityOfParentalPhasedLoci: quantityOfParentalPhasedLoci,
						}

						newOffspringGenomePairTraitInfo.NeuralNetworkAnalysis = newOffspringGenomePairTraitInfo_NeuralNetwork
					}
				}

				anyRulesExist, rulesAnalysisExists, quantityOfRulesTested, quantityOfLociKnown, offspringProbabilityOfPassingRulesMap, offspringOutcomeProbabilitiesMap, err := GetOffspringDiscreteTraitAnalysis_Rules(traitObject, person1LocusValuesMap, person2LocusValuesMap)
				if (err != nil) { return err }
				if (anyRulesExist == true){

					newOffspringGenomePairTraitInfo.RulesExist = true

					if (rulesAnalysisExists == true){
						newOffspringGenomePairTraitInfo.RulesAnalysisExists = true

						newOffspringGenomePairTraitInfo_Rules := geneticAnalysis.OffspringGenomePairDiscreteTraitInfo_Rules{
							QuantityOfRulesTested: quantityOfRulesTested,
							QuantityOfLociKnown: quantityOfLociKnown,
							ProbabilityOfPassingRulesMap: offspringProbabilityOfPassingRulesMap,
							OffspringOutcomeProbabilitiesMap: offspringOutcomeProbabilitiesMap,
						}

						newOffspringGenomePairTraitInfo.RulesAnalysis = newOffspringGenomePairTraitInfo_Rules
					}
				}

				genomePairIdentifier := helpers.JoinTwo16ByteArrays(person1GenomeIdentifier, person2GenomeIdentifier)

				offspringTraitInfoMap[genomePairIdentifier] = newOffspringGenomePairTraitInfo

				return nil
			}

			err = addGenomePairTraitInfoToOffspringMap(pair1Person1GenomeIdentifier, pair1Person2GenomeIdentifier)
			if (err != nil) { return false, "", err }

			if (genomePair2Exists == true){

				err := addGenomePairTraitInfoToOffspringMap(pair2Person1GenomeIdentifier, pair2Person2GenomeIdentifier)
				if (err != nil) { return false, "", err }
			}

			newOffspringTraitInfoObject := geneticAnalysis.OffspringDiscreteTraitInfo{
				TraitInfoMap: offspringTraitInfoMap,
			}

			if (len(offspringTraitInfoMap) >= 2){

				// We check for conflicts
				// Conflicts are only possible if two genome pairs exist with information about the trait

				checkIfConflictExists := func()(bool, error){

					// We check for conflicts between each genome pair's outcome scores and trait rules maps

					genomePairTraitInfoObject := geneticAnalysis.OffspringGenomePairDiscreteTraitInfo{}

					firstItemReached := false

					for _, currentGenomePairTraitInfoObject := range offspringTraitInfoMap{

						if (firstItemReached == false){
							genomePairTraitInfoObject = currentGenomePairTraitInfoObject
							firstItemReached = true
							continue
						}

						areEqual := reflect.DeepEqual(genomePairTraitInfoObject, currentGenomePairTraitInfoObject)
						if (areEqual == false){
							return true, nil
						}
					}

					return false, nil
				}

				conflictExists, err := checkIfConflictExists()
				if (err != nil) { return false, "", err }

				newOffspringTraitInfoObject.ConflictExists = conflictExists
			}

			offspringDiscreteTraitsMap[traitName] = newOffspringTraitInfoObject

		} else {

		// traitIsDiscreteOrNumeric == "Numeric"

			// This map stores the trait info for each genome pair
			// Map Structure: Genome Pair Identifier -> OffspringGenomePairNumericTraitInfo
			offspringTraitInfoMap := make(map[[32]byte]geneticAnalysis.OffspringGenomePairNumericTraitInfo)

			// This will add the offspring trait information for the provided genome pair to the offspringTraitInfoMap
			addGenomePairTraitInfoToOffspringMap := func(person1GenomeIdentifier [16]byte, person2GenomeIdentifier [16]byte)error{

				person1LocusValuesMap, exists := person1GenomesMap[person1GenomeIdentifier]
				if (exists == false){
					return errors.New("addGenomePairTraitInfoToOffspringMap called with unknown person1GenomeIdentifier.")
				}

				person2LocusValuesMap, exists := person2GenomesMap[person2GenomeIdentifier]
				if (exists == false){
					return errors.New("addGenomePairTraitInfoToOffspringMap called with unknown person2GenomeIdentifier.")
				}

				neuralNetworkExists, neuralNetworkAnalysisExists, averageOutcome, predictionConfidenceRangesMap, sampleOffspringOutcomesList, quantityOfLociTested, quantityOfParentalPhasedLoci, err := GetOffspringNumericTraitAnalysis(traitObject, person1LocusValuesMap, person2LocusValuesMap)
				if (err != nil) { return err }
				if (neuralNetworkExists == false){
					// Predictions are not possible for this trait
					return nil
				}
				if (neuralNetworkAnalysisExists == false){
					// No locations for this trait exists in which both user's genomes contain information
					return nil
				}

				newOffspringGenomePairTraitInfo := geneticAnalysis.OffspringGenomePairNumericTraitInfo{
					OffspringAverageOutcome: averageOutcome,
					PredictionConfidenceRangesMap: predictionConfidenceRangesMap,
					QuantityOfParentalPhasedLoci: quantityOfParentalPhasedLoci,
					QuantityOfLociKnown: quantityOfLociTested,
					SampleOffspringOutcomesList: sampleOffspringOutcomesList,
				}

				genomePairIdentifier := helpers.JoinTwo16ByteArrays(person1GenomeIdentifier, person2GenomeIdentifier)

				offspringTraitInfoMap[genomePairIdentifier] = newOffspringGenomePairTraitInfo

				return nil
			}

			err = addGenomePairTraitInfoToOffspringMap(pair1Person1GenomeIdentifier, pair1Person2GenomeIdentifier)
			if (err != nil) { return false, "", err }

			if (genomePair2Exists == true){

				err := addGenomePairTraitInfoToOffspringMap(pair2Person1GenomeIdentifier, pair2Person2GenomeIdentifier)
				if (err != nil) { return false, "", err }
			}

			newOffspringTraitInfoObject := geneticAnalysis.OffspringNumericTraitInfo{
				TraitInfoMap: offspringTraitInfoMap,
			}

			if (len(offspringTraitInfoMap) >= 2){

				// We check for conflicts
				// Conflicts are only possible if two genome pairs exist with information about the trait

				checkIfConflictExists := func()(bool, error){

					// We check for conflicts between each genome pair's outcome scores and trait rules maps

					genomePairTraitInfoObject := geneticAnalysis.OffspringGenomePairNumericTraitInfo{}

					firstItemReached := false

					for _, currentGenomePairTraitInfoObject := range offspringTraitInfoMap{

						if (firstItemReached == false){
							genomePairTraitInfoObject = currentGenomePairTraitInfoObject
							firstItemReached = true
							continue
						}

						areEqual := reflect.DeepEqual(genomePairTraitInfoObject, currentGenomePairTraitInfoObject)
						if (areEqual == false){
							return true, nil
						}
					}

					return false, nil
				}

				conflictExists, err := checkIfConflictExists()
				if (err != nil) { return false, "", err }

				newOffspringTraitInfoObject.ConflictExists = conflictExists
			}

			offspringNumericTraitsMap[traitName] = newOffspringTraitInfoObject
		}
	}

	newCoupleAnalysis.DiscreteTraitsMap = offspringDiscreteTraitsMap
	newCoupleAnalysis.NumericTraitsMap = offspringNumericTraitsMap

	analysisBytes, err := encoding.EncodeMessagePackBytes(newCoupleAnalysis)
	if (err != nil) { return false, "", err }

	analysisString := string(analysisBytes)

	return true, analysisString, nil
}


// We also use this function when calculating offspring probabilities between users in viewProfileGui.go
//Outputs:
//	-bool: Probability offspring has disease is known
//	-int: Percentage probability offspring has disease (0-100)
//	-bool: Probability offspring has variant is known
//	-int: Percentage probability offspring has variant (0-100)
//	-error
func GetOffspringMonogenicDiseaseProbabilities(dominantOrRecessive string, person1ProbabilityIsKnown bool, person1WillPassVariantPercentageProbability int, person2ProbabilityIsKnown bool, person2WillPassVariantPercentageProbability int)(bool, int, bool, int, error){

	if (dominantOrRecessive != "Dominant" && dominantOrRecessive != "Recessive"){
		return false, 0, false, 0, errors.New("GetOffspringMonogenicDiseaseProbabilities called with invalid dominantOrRecessive: " + dominantOrRecessive)
	}
	if (person1ProbabilityIsKnown == false && person2ProbabilityIsKnown == false){
		return false, 0, false, 0, nil
	}

	if (person1ProbabilityIsKnown == true){
		if (person1WillPassVariantPercentageProbability < 0 || person1WillPassVariantPercentageProbability > 100){
			return false, 0, false, 0, errors.New("GetOffspringMonogenicDiseaseProbabilities called with invalid person1WillPassVariantProbability")
		}
	}
	if (person2ProbabilityIsKnown == true){
		if (person2WillPassVariantPercentageProbability < 0 || person2WillPassVariantPercentageProbability > 100){
			return false, 0, false, 0, errors.New("GetOffspringMonogenicDiseaseProbabilities called with invalid person2WillPassVariantProbability")
		}
	}
	if (person1ProbabilityIsKnown == false || person2ProbabilityIsKnown == false){
		// Only 1 of the person's probabilities are known

		getPersonWhoHasVariantProbabilityOfPassingIt := func()int{
			if (person1ProbabilityIsKnown == true){
				return person1WillPassVariantPercentageProbability
			}
			return person2WillPassVariantPercentageProbability
		}

		personWhoHasVariantProbabilityOfPassingIt := getPersonWhoHasVariantProbabilityOfPassingIt()

		if (personWhoHasVariantProbabilityOfPassingIt == 100){
			if (dominantOrRecessive == "Dominant"){
				// We know the offspring will have the disease and will have a variant
				return true, 100, true, 100, nil
			}
			//dominantOrRecessive == "Recessive"
			// We don't know if the offspring will have the disease, but we know they will have a variant
			return false, 0, true, 100, nil
		}
		if (personWhoHasVariantProbabilityOfPassingIt == 0){

			if (dominantOrRecessive == "Recessive"){
				// We know the offspring will not have the disease, but we don't know if they will have a variant
				return true, 0, false, 0, nil
			}
		}

		// We don't know the probabilities that the offspring will have the disease or if they will have a variant
		return false, 0, false, 0, nil
	}

	person1WillPassVariantProbability := float64(person1WillPassVariantPercentageProbability)/100
	person2WillPassVariantProbability := float64(person2WillPassVariantPercentageProbability)/100

	// The probability offspring has a variant = the probability that either parent passes a variant (inclusive or)
	// We find the probability of the offspring having a monogenic disease variant as follows:
	// P(A U B) = P(A) + P(B) - P(A ∩ B)
	// (Probability of person 1 passing a variant) + (Probability of person 2 passing a variant) - (Probability of offspring having disease)
	// A person with a variant may have the disease, or just be a carrier.
	probabilityOffspringHasVariant := person1WillPassVariantProbability + person2WillPassVariantProbability - (person1WillPassVariantProbability * person2WillPassVariantProbability)

	if (dominantOrRecessive == "Dominant"){

		// The probability of having the monogenic disease is the same as the probability of having a variant

		percentageProbabilityOffspringHasVariant := int(probabilityOffspringHasVariant * 100)

		return true, percentageProbabilityOffspringHasVariant, true, percentageProbabilityOffspringHasVariant, nil
	}

	// We find the probability of the offspring having the mongenic disease as follows:
	// P(A and B) = P(A) * P(B)
	// (Probability of person 1 Passing a variant) * (Probability of person 2 passing a variant)
	probabilityOffspringHasDisease := person1WillPassVariantProbability * person2WillPassVariantProbability

	percentageProbabilityOffspringHasDisease := probabilityOffspringHasDisease * 100
	percentageProbabilityOffspringHasVariant := probabilityOffspringHasVariant * 100

	// This conversion remove any digits after the radix point
	// This will not result in any false 0% values, an example being 0.9% becoming 0%
	// This is because the lowest non-zero probability a person can have for passing a variant is 50%
	// Thus, the lowest non-zero probability of an offspring having a disease is 25%
	percentageProbabilityOffspringHasDiseaseInt := int(percentageProbabilityOffspringHasDisease)
	percentageProbabilityOffspringHasVariantInt := int(percentageProbabilityOffspringHasVariant)

	return true, percentageProbabilityOffspringHasDiseaseInt, true, percentageProbabilityOffspringHasVariantInt, nil
}

//Outputs:
//	-bool: A neural network exists for this trait
//	-bool: Any loci tested (if false, no offspring polygenic disease analysis is known)
//	-int: Offspring Average Risk Score (Value between 0-10)
//	-map[int]float64: Prediction accuracy ranges map
//		-Map Structure: Probability prediction is accurate (X) -> Distance from prediction that must be travelled in both directions to
//			create a range in which the true value will fall into, X% of the time
//	-[]int: Sample offspring risks scores list
//	-int: Quantity of loci known
//	-int: Quantity of parental phased loci
//	-error
func GetOffspringPolygenicDiseaseAnalysis(diseaseObject polygenicDiseases.PolygenicDisease, person1LocusValuesMap map[int64]locusValue.LocusValue, person2LocusValuesMap map[int64]locusValue.LocusValue)(bool, bool, int, map[int]float64, []int, int, int, error){

	diseaseName := diseaseObject.DiseaseName

	modelExists := trainedPredictionModels.CheckIfAttributeNeuralNetworkExists(diseaseName)
	if (modelExists == false){
		// Prediction is not possible for this trait
		return false, false, 0, nil, nil, 0, 0, nil
	}

	if (len(person1LocusValuesMap) == 0){
		return true, false, 0, nil, nil, 0, 0, nil
	}
	if (len(person2LocusValuesMap) == 0){
		return true, false, 0, nil, nil, 0, 0, nil
	}

	diseaseLociList := diseaseObject.LociList

	// First we count up the quantity of parental phased loci
	// We only count the quantity of phased loci for loci which are known for both parents

	quantityOfParentalPhasedLoci := 0

	for _, rsID := range diseaseLociList{

		person1LocusValue, exists := person1LocusValuesMap[rsID]
		if (exists == false){
			continue
		}

		person2LocusValue, exists := person2LocusValuesMap[rsID]
		if (exists == false){
			continue
		}

		person1LocusIsPhased := person1LocusValue.LocusIsPhased
		if (person1LocusIsPhased == true){
			quantityOfParentalPhasedLoci += 1
		}

		person2LocusIsPhased := person2LocusValue.LocusIsPhased
		if (person2LocusIsPhased == true){
			quantityOfParentalPhasedLoci += 1
		}
	}

	// We create 100 prospective offspring genomes.

	anyLocusValueExists, prospectiveOffspringGenomesList, err := getProspectiveOffspringGenomesList(diseaseLociList, person1LocusValuesMap, person2LocusValuesMap)
	if (err != nil) { return false, false, 0, nil, nil, 0, 0, err }
	if (anyLocusValueExists == false){
		return true, false, 0, nil, nil, 0, 0, nil
	}

	// A list of predicted risk scores for each offspring
	predictedRiskScoresList := make([]int, 0)

	accuracyRangesMap := make(map[int]float64)
	quantityOfLociTested := 0

	for index, offspringGenomeMap := range prospectiveOffspringGenomesList{

		neuralNetworkExists, predictionIsKnown, predictedRiskScore, predictionAccuracyRangesMap, currentQuantityOfLociTested, _, err := createPersonGeneticAnalysis.GetPersonGenomePolygenicDiseaseAnalysis(diseaseObject, offspringGenomeMap, false)
		if (err != nil){ return false, false, 0, nil, nil, 0, 0, err }
		if (neuralNetworkExists == false){
			return false, false, 0, nil, nil, 0, 0, errors.New("GetGenomeNumericTraitAnalysis claiming that neural network doesn't exist when we already checked.")
		}
		if (predictionIsKnown == false){
			return false, false, 0, nil, nil, 0, 0, errors.New("GetGenomeNumericTraitAnalysis claiming that prediction is impossible when we already know at least 1 locus value exists for trait.")
		}

		predictedRiskScoresList = append(predictedRiskScoresList, predictedRiskScore)

		if (index == 0){
			// These values should be the same for each predicted offspring
			accuracyRangesMap = predictionAccuracyRangesMap
			quantityOfLociTested = currentQuantityOfLociTested
		}
	}

	// We calculate the average predicted risk score

	outcomesSum := 0

	for _, predictedRiskScore := range predictedRiskScoresList{
		outcomesSum += predictedRiskScore
	}

	averageRiskScore := outcomesSum/100

	return true, true, averageRiskScore, accuracyRangesMap, predictedRiskScoresList, quantityOfLociTested, quantityOfParentalPhasedLoci, nil
}


//Outputs:
//	-bool: A neural network exists for this trait
//	-bool: Analysis exists (at least 1 locus exists for this analysis from both people's genomes
//	-map[string]int: Outcome probabilities map
//		Map Structure: Outcome Name -> Offspring probability of outcome
//	-int: Average prediction confidence (the average prediction confidence for all prospective offspring)
//	-int: Quantity of loci tested
//	-int: Quantity of parental phased loci
//	-error
func GetOffspringDiscreteTraitAnalysis_NeuralNetwork(traitObject traits.Trait, person1LocusValuesMap map[int64]locusValue.LocusValue, person2LocusValuesMap map[int64]locusValue.LocusValue)(bool, bool, map[string]int, int, int, int, error){

	traitName := traitObject.TraitName

	traitIsDiscreteOrNumeric := traitObject.DiscreteOrNumeric
	if (traitIsDiscreteOrNumeric != "Discrete"){
		return false, false, nil, 0, 0, 0, errors.New("GetOffspringDiscreteTraitAnalysis_NeuralNetwork called with non-discrete trait.")
	}

	modelExists := trainedPredictionModels.CheckIfAttributeNeuralNetworkExists(traitName)
	if (modelExists == false){
		// Neural network prediction is not possible for this trait
		return false, false, nil, 0, 0, 0, nil
	}

	traitLociList := traitObject.LociList

	// First we count up the quantity of parental phased loci
	// We only count the quantity of phased loci for loci which are known for both parents

	quantityOfParentalPhasedLoci := 0

	for _, rsID := range traitLociList{

		person1LocusValue, exists := person1LocusValuesMap[rsID]
		if (exists == false){
			continue
		}

		person2LocusValue, exists := person2LocusValuesMap[rsID]
		if (exists == false){
			continue
		}

		person1LocusIsPhased := person1LocusValue.LocusIsPhased
		if (person1LocusIsPhased == true){
			quantityOfParentalPhasedLoci += 1
		}

		person2LocusIsPhased := person2LocusValue.LocusIsPhased
		if (person2LocusIsPhased == true){
			quantityOfParentalPhasedLoci += 1
		}
	}

	// Next, we create 100 prospective offspring genomes.

	anyLocusValueExists, prospectiveOffspringGenomesList, err := getProspectiveOffspringGenomesList(traitLociList, person1LocusValuesMap, person2LocusValuesMap)
	if (err != nil) { return false, false, nil, 0, 0, 0, err }
	if (anyLocusValueExists == false){
		return true, false, nil, 0, 0, 0, nil
	}

	// Map Structure: Outcome Name -> Probability of outcome coming true
	// Because we are summing from 100 offspring, the count of outcomes is the same as the probability of an offspring having the outcome
	outcomeCountsMap := make(map[string]int)

	// This is a sum of each prediction's confidence
	predictionConfidencesSum := 0

	quantityOfLociTested := 0

	for index, offspringGenomeMap := range prospectiveOffspringGenomesList{

		neuralNetworkExists, predictionIsKnown, predictedOutcome, predictionConfidence, currentQuantityOfLociTested, _, err := createPersonGeneticAnalysis.GetGenomeDiscreteTraitAnalysis_NeuralNetwork(traitObject, offspringGenomeMap, false)
		if (err != nil){ return false, false, nil, 0, 0, 0, err }
		if (neuralNetworkExists == false){
			return false, false, nil, 0, 0, 0, errors.New("GetGenomeTraitAnalysis_NeuralNetwork claiming that neural network doesn't exist when we already checked.")
		}
		if (predictionIsKnown == false){
			return false, false, nil, 0, 0, 0, errors.New("GetGenomeTraitAnalysis_NeuralNetwork claiming that prediction is impossible when we already know at least 1 locus value exists for trait.")
		}

		outcomeCountsMap[predictedOutcome] += 1
		predictionConfidencesSum += predictionConfidence

		if (index == 0){
			// This value should be the same for each predicted offspring
			quantityOfLociTested = currentQuantityOfLociTested
		}
	}

	averagePredictionConfidence := predictionConfidencesSum/100

	return true, true, outcomeCountsMap, averagePredictionConfidence, quantityOfLociTested, quantityOfParentalPhasedLoci, nil
}

//Outputs:
//	-bool: Any rules exist (if false, rule-based prediction is not possible for this trait)
//	-bool: Rule-based analysis exists (if false, no offspring trait information is known, or there is an outcome tie for one of the offspring)
//	-int: Quantity of rules tested
//	-int: Quantity of loci known
//	-map[[3]byte]int: Offspring probability of passing rules map
//		Map Structure: Rule identifier -> Offspring probability of passing rule (1-100)
//		If a rule entry doesn't exist, we don't know the passes-rule probability for any of the offspring
//	-map[string]int: Offspring outcome probabilities map
//		Map Structure: Outcome Name -> Offspring probability of outcome (0-100)
//	-error
func GetOffspringDiscreteTraitAnalysis_Rules(traitObject traits.Trait, person1LocusValuesMap map[int64]locusValue.LocusValue, person2LocusValuesMap map[int64]locusValue.LocusValue)(bool, bool, int, int, map[[3]byte]int, map[string]int, error){

	traitRulesList := traitObject.RulesList

	if (len(traitRulesList) == 0){
		return false, false, 0, 0, nil, nil, nil
	}

	if (len(person1LocusValuesMap) == 0){
		return true, false, 0, 0, nil, nil, nil
	}
	if (len(person2LocusValuesMap) == 0){
		return true, false, 0, 0, nil, nil, nil
	}

	// First, we create 100 prospective offspring genomes.

	traitLociList_Rules := traitObject.LociList_Rules

	anyLocusValueExists, prospectiveOffspringGenomesList, err := getProspectiveOffspringGenomesList(traitLociList_Rules, person1LocusValuesMap, person2LocusValuesMap)
	if (err != nil) { return false, false, 0, 0, nil, nil, err }
	if (anyLocusValueExists == false){
		return true, false, 0, 0, nil, nil, nil
	}

	// Map Structure: Rule Identifier -> Number of offspring who pass the rule (out of 100 prospective offspring)
	// Because there are 100 offspring, this also represents the percentage probability that an offspring will pass the rule
	offspringPassesRulesCountMap := make(map[[3]byte]int)

	// This map stores the quantity of offspring who have each outcome
	// The probability an offspring will have this outcome is the same as the
	//		quantity of offspring who have this outcome in our set of 100 randomly generated offspring
	// Map structure: Outcome name -> quantity of offspring who have this outcome
	outcomeCountsMap := make(map[string]int)

	quantityOfLociKnown := 0

	for index, offspringGenomeMap := range prospectiveOffspringGenomesList{

		// Now we get outcome prediction for prospective offspring

		anyRulesExist, quantityOfRulesTested, currentQuantityOfLociKnown, offspringPassesRulesMap, predictionOutcomeIsKnown, predictedOutcome, err := createPersonGeneticAnalysis.GetGenomeDiscreteTraitAnalysis_Rules(traitObject, offspringGenomeMap, false)
		if (err != nil) { return false, false, 0, 0, nil, nil, err }
		if (anyRulesExist == false){
			return false, false, 0, 0, nil, nil, errors.New("GetGenomeTraitAnalysis_Rules returning noRulesExists when we already checked and trait rules do in-fact exist.")
		}
		if (quantityOfRulesTested == 0){
			// This will be the same for each of the 100 generated offspring
			// No analysis is possible.
			return true, false, 0, currentQuantityOfLociKnown, nil, nil, nil
		}
		if (index == 0){
			// currentQuantityOfLociKnown will be the same for each prospective offspring
			quantityOfLociKnown = currentQuantityOfLociKnown
		}

		for ruleIdentifier, genomePassesRule := range offspringPassesRulesMap{
			if (genomePassesRule == true){
				offspringPassesRulesCountMap[ruleIdentifier] += 1
			}
		}

		if (predictionOutcomeIsKnown == false){
			// There was a tie between outcomes for this offspring
			// We can't predict anything about this trait for this couple using rules
			// This is why we need to create rules which make it unlikely for a tie between outcomes to occur.
			return true, false, 0, 0, nil, nil, nil
		}

		outcomeCountsMap[predictedOutcome] += 1
	}

	quantityOfRulesTested := len(offspringPassesRulesCountMap)

	return true, true, quantityOfRulesTested, quantityOfLociKnown, offspringPassesRulesCountMap, outcomeCountsMap, nil
}


//Outputs:
//	-bool: A neural network exists for this trait
//	-bool: Analysis exists (at least 1 locus exists for this analysis from both people's genomes
//	-float64: Average outcome for offspring
//	-map[int]float64: Prediction accuracy ranges map
//		-Map Structure: Probability prediction is accurate (X) -> Distance from predictoin that must be travelled in both directions to
//			create a range in which the true value will fall into, X% of the time
//	-[]float64: A list of 100 offspring outcomes
//	-int: Quantity of loci known
//	-int: Quantity of parental phased loci
//	-error
func GetOffspringNumericTraitAnalysis(traitObject traits.Trait, person1LocusValuesMap map[int64]locusValue.LocusValue, person2LocusValuesMap map[int64]locusValue.LocusValue)(bool, bool, float64, map[int]float64, []float64, int, int, error){

	traitName := traitObject.TraitName

	traitIsDiscreteOrNumeric := traitObject.DiscreteOrNumeric
	if (traitIsDiscreteOrNumeric != "Numeric"){
		return false, false, 0, nil, nil, 0, 0, errors.New("GetOffspringNumericTraitAnalysis called with non-numeric trait.")
	}

	modelExists := trainedPredictionModels.CheckIfAttributeNeuralNetworkExists(traitName)
	if (modelExists == false){
		// Prediction is not possible for this trait
		return false, false, 0, nil, nil, 0, 0, nil
	}

	traitLociList := traitObject.LociList

	// First we count up the quantity of parental phased loci
	// We only count the quantity of phased loci for loci which are known for both parents

	quantityOfParentalPhasedLoci := 0

	for _, rsID := range traitLociList{

		person1LocusValue, exists := person1LocusValuesMap[rsID]
		if (exists == false){
			continue
		}

		person2LocusValue, exists := person2LocusValuesMap[rsID]
		if (exists == false){
			continue
		}

		person1LocusIsPhased := person1LocusValue.LocusIsPhased
		if (person1LocusIsPhased == true){
			quantityOfParentalPhasedLoci += 1
		}

		person2LocusIsPhased := person2LocusValue.LocusIsPhased
		if (person2LocusIsPhased == true){
			quantityOfParentalPhasedLoci += 1
		}
	}

	// Next, we create 100 prospective offspring genomes.

	anyLocusValueExists, prospectiveOffspringGenomesList, err := getProspectiveOffspringGenomesList(traitLociList, person1LocusValuesMap, person2LocusValuesMap)
	if (err != nil) { return false, false, 0, nil, nil, 0, 0, err }
	if (anyLocusValueExists == false){
		return true, false, 0, nil, nil, 0, 0, nil
	}

	// A list of predicted outcomes for each offspring
	predictedOutcomesList := make([]float64, 0)

	accuracyRangesMap := make(map[int]float64)
	quantityOfLociTested := 0

	for index, offspringGenomeMap := range prospectiveOffspringGenomesList{

		neuralNetworkExists, predictionIsKnown, predictedOutcome, predictionAccuracyRangesMap, currentQuantityOfLociTested, _, err := createPersonGeneticAnalysis.GetGenomeNumericTraitAnalysis(traitObject, offspringGenomeMap, false)
		if (err != nil){ return false, false, 0, nil, nil, 0, 0, err }
		if (neuralNetworkExists == false){
			return false, false, 0, nil, nil, 0, 0, errors.New("GetGenomeNumericTraitAnalysis claiming that neural network doesn't exist when we already checked.")
		}
		if (predictionIsKnown == false){
			return false, false, 0, nil, nil, 0, 0, errors.New("GetGenomeNumericTraitAnalysis claiming that prediction is impossible when we already know at least 1 locus value exists for trait.")
		}

		predictedOutcomesList = append(predictedOutcomesList, predictedOutcome)

		if (index == 0){
			// These values should be the same for each predicted offspring
			accuracyRangesMap = predictionAccuracyRangesMap
			quantityOfLociTested = currentQuantityOfLociTested
		}
	}

	// We calculate the average predicted outcome

	outcomesSum := float64(0)

	for _, predictedOutcome := range predictedOutcomesList{
		outcomesSum += predictedOutcome
	}

	averageOutcome := outcomesSum/100

	return true, true, averageOutcome, accuracyRangesMap, predictedOutcomesList, quantityOfLociTested, quantityOfParentalPhasedLoci, nil
}

// This function will return a list of 100 prospective offspring genomes
// Each genome represents an equal-probability offspring genome from both people's genomes
// This function takes into account the effects of genetic linkage
// Any locations which do not exist in both people's genomes will not be included
//
// TODO: The user should be able to choose how many prospective offspring to create in the settings.
//		More offspring will take longer, but will yield a more accurate trait analysis.
//
//Outputs:
//	-bool: Any locus value exists between both users
//	-[]map[int64]locusValue.LocusValue
//	-error
func getProspectiveOffspringGenomesList(lociList []int64, person1LociMap map[int64]locusValue.LocusValue, person2LociMap map[int64]locusValue.LocusValue)(bool, []map[int64]locusValue.LocusValue, error){

	// -We use randomness to generate the offspring genomes
	// -We want the results to be the same for each pair of people each time, so we have to seed our randomness generator
	// -This is necessary so that two people's analysis results do not change every time
	// -Instead, the same 2 people will produce the exact same result every time
	pseudorandomNumberGenerator := mathRand.New(mathRand.NewPCG(1, 2))

	//Outputs:
	// -[]int64: A list of random breakpoints for this chromosome that are statistically accurate
	// -error
	getRandomChromosomeBreakpoints := func(chromosome int)([]int64, error){

		getChromosomeLength := func()(int64, error){

			// Approximate number of base pairs in each chromosome taken from: https://www.ncbi.nlm.nih.gov/books/NBK557784/

			switch chromosome{

				case 1:{
					return 249000000, nil
				}
				case 2:{
					return 243000000, nil
				}
				case 3:{
					return 200000000, nil
				}
				case 4:{
					return 192000000, nil
				}
				case 5:{
					return 181000000, nil
				}
				case 6:{
					return 170000000, nil
				}
				case 7:{
					return 158000000, nil
				}
				case 8:{
					return 146000000, nil
				}
				case 9:{
					return 140000000, nil
				}
				case 10:{
					return 135000000, nil
				}
				case 11:{
					return 135000000, nil
				}
				case 12:{
					return 132000000, nil
				}
				case 13:{
					return 114000000, nil
				}
				case 14:{
					return 106000000, nil
				}
				case 15:{
					return 100000000, nil
				}
				case 16:{
					return 89000000, nil
				}
				case 17:{
					return 79000000, nil
				}
				case 18:{
					return 76000000, nil
				}
				case 19:{
					return 64000000, nil
				}
				case 20:{
					return 62000000, nil
				}
				case 21:{
					return 47000000, nil
				}
				case 22:{
					return 50000000, nil
				}
			}

			chromosomeString := helpers.ConvertIntToString(chromosome)
			return 0, errors.New("getRandomChromosomeBreakpoints called with invalid chromosome: " + chromosomeString)
		}

		chromosomeLength, err := getChromosomeLength()
		if (err != nil) { return nil, err }

		listOfRandomBreakpoints := make([]int64, 0)

		// TODO: Take into account different recombination rate for each chromosome
		// TODO: There are also breakpoint hotspots which we need to account for
		// TODO: I read somewhere that recombination break points are less likely to occur within genes,
		//		meaning they are more likely to occur at the gene boundaries (codons)

		// We step by 1,000,000 each time
		// It would be more realistic if we did it in 1 integer increments, but it would be slower
		for position := int64(0); position < chromosomeLength; position += 1_000_000{

			//From Wikipedia:
			// A centimorgan (abbreviated cM) is a unit for measuring genetic linkage.
			// It is defined as the distance between chromosome positions (loci) for which the expected
			//		average number of intervening chromosomal crossovers in a single generation is 0.01.
			// One centimorgan corresponds to about 1 million base pairs in humans on average
			//
			// A chromosomal crossover == recombination breakpoint
			//
			// For every 1,000,000 base pairs, there is a 0.01 probability that there is a breakpoint

			randomFloat := pseudorandomNumberGenerator.Float64()
			if (randomFloat <= 0.01){
				// This has a 0.01, or 1% probability of being true
				listOfRandomBreakpoints = append(listOfRandomBreakpoints, position)
			}
		}

		return listOfRandomBreakpoints, nil
	}

	// Map Structure: rsID -> Locus Value
	offspringGenomesList := make([]map[int64]locusValue.LocusValue, 0)

	for i:=0; i < 100; i++{

		// This map stores the chromosome breakpoints for person1
		// Map Structure: Chromosome -> List of breakpoints
		person1ChromosomeBreakpointsMap := make(map[int][]int64)

		// This map stores the chromosome breakpoints for person2
		// Map Structure: Chromosome -> List of breakpoints
		person2ChromosomeBreakpointsMap := make(map[int][]int64)

		// This stores the locus values for this prospective offspring
		// Map Structure: rsID -> Locus Value
		prospectiveOffspringGenome := make(map[int64]locusValue.LocusValue)

		for _, rsID := range lociList{

			//Outputs:
			//	-bool: Allele is known
			//	-string: Locus base
			//	-error
			getPersonAllele := func(personLociMap map[int64]locusValue.LocusValue, personBreakpointsMap map[int][]int64)(bool, string, error){

				personLocusValue, exists := personLociMap[rsID]
				if (exists == false){
					return false, "", nil
				}

				personLocusBase1 := personLocusValue.Base1Value
				personLocusBase2 := personLocusValue.Base2Value
				personLocusIsPhased := personLocusValue.LocusIsPhased

				if (personLocusBase1 == personLocusBase2){
					// Phase doesn't matter
					return true, personLocusBase1, nil
				}

				if (personLocusIsPhased == false){
					// Breakpoints are unnecessary
					// We either choose base 1 or 2
					randomInt := pseudorandomNumberGenerator.IntN(2)
					if (randomInt == 1){
						return true, personLocusBase1, nil
					}
					return true, personLocusBase2, nil
				}

				// We have a phased locus
				// We figure out which allele to use by seeing which allele gets inherited from our random breakpoints list
				// We figure out the chromosome and position of this locus

				locusMetadataExists, locusMetadataObject, err := locusMetadata.GetLocusMetadata(rsID)
				if (err != nil) { return false, "", err }
				if (locusMetadataExists == false){
					rsIDString := helpers.ConvertInt64ToString(rsID)
					return false, "", errors.New("getProspectiveOffspringGenomesList called with unknown rsID: " + rsIDString)
				}

				locusPosition := locusMetadataObject.Position
				locusChromosome := locusMetadataObject.Chromosome

				getPersonChromosomeBreakpointsList := func()([]int64, error){

					breakpointsList, exists := personBreakpointsMap[locusChromosome]
					if (exists == true){
						return breakpointsList, nil
					}

					// We have to create a new breakpoints list

					newBreakpointsList, err := getRandomChromosomeBreakpoints(locusChromosome)
					if (err != nil) { return nil, err }

					personBreakpointsMap[locusChromosome] = newBreakpointsList

					return newBreakpointsList, nil
				}

				personBreakpointsList, err := getPersonChromosomeBreakpointsList()
				if (err != nil) { return false, "", err }

				getLocusListIndex := func()int{

					for index, breakpointPosition := range personBreakpointsList{

						if (int64(locusPosition) <= breakpointPosition){

							return index
						}
					}

					index := len(personBreakpointsList)

					// This is reached if the final breakpoint in the list is less than the locus's position, or if there were no breakpoints

					return index
				}

				locusListIndex := getLocusListIndex()

				if (locusListIndex%2 == 0){
					return true, personLocusBase1, nil
				}

				return true, personLocusBase2, nil
			}

			person1AlleleIsKnown, person1Allele, err := getPersonAllele(person1LociMap, person1ChromosomeBreakpointsMap)
			if (err != nil) { return false, nil, err }
			if (person1AlleleIsKnown == false){
				continue
			}

			person2AlleleIsKnown, person2Allele, err := getPersonAllele(person2LociMap, person2ChromosomeBreakpointsMap)
			if (err != nil) { return false, nil, err }
			if (person2AlleleIsKnown == false){
				continue
			}

			offspringLocusValue := locusValue.LocusValue{
				Base1Value: person1Allele,
				Base2Value: person2Allele,
				LocusIsPhased: true,
			}

			prospectiveOffspringGenome[rsID] = offspringLocusValue
		}

		if (len(prospectiveOffspringGenome) == 0){
			// We don't have any locations at which both people's genomes contain a locus value.
			return false, nil, nil
		}

		offspringGenomesList = append(offspringGenomesList, prospectiveOffspringGenome)	
	}

	return true, offspringGenomesList, nil
}