public async Task <PlanElementCandidate> GenerateAsync(DecisionArray decisionArray, Plan plan, int whichMeal, PlanElementIteratorParams iterParams)
{
var test = true;
//choose type based on preferences
var type = (plan.PlanForm.FoodPreference == FoodPreference.OnlyRestaurant || (plan.PlanForm.FoodPreference == FoodPreference.Mixed && whichMeal == 2)) ? GooglePlaceTypeCategory.Restaurant : GooglePlaceTypeCategory.Food;
var candidates = new List <PlanElementCandidate>();
var decisionRows = new List <DecisionRow>();
var googleNearbyFoodInput = _googlePlaceNearbySearchInputFactory.Create(iterParams.CurrentLocation, type);
var nearbyFoodResults = await _googlePlaceNearbySearchApiClient.GetAsync(googleNearbyFoodInput);
int counter = 1;
foreach (var nr in nearbyFoodResults.results)
{
var details = await _googlePlaceDetailsApiClient.GetAsync(_googlePlaceDetailsInputFactory.CreateAllUseful(nr.place_id));
if (details.IsOk)
{
var candidate = new PlanElementCandidate(details.Result.name, details.Result.place_id, details.Result.formatted_address, details.Result.geometry.location, details.Result.opening_hours, details.Result.types, details.Result.rating, details.Result.price_level, details.Result.user_ratings_total);
if (test)
{
return(candidate);
}
candidates.Add(candidate);
}
++counter;
if (counter > 10 && candidates.Count > 5)
{
break;
}
}
var iter = 1;
foreach (var candidate in candidates)
{
decisionRows.Add(_decisionRowFactory.Create(candidate, iter, iterParams.CurrentLocation));
++iter;
}
var minVector = DecisionArray.GetMinVector(decisionRows);
var maxVector = DecisionArray.GetMaxVector(decisionRows);
foreach (var decisionRow in decisionRows)
{
decisionRow.NormalizedScore = _sawMethod.CalculateNormalizedScore(SawNormalizationMethod.LinearFirstType, decisionArray.WeightVector, decisionRow.DecisionValues, minVector, maxVector);
}
var result = decisionRows.OrderByDescending(x => x.NormalizedScore).FirstOrDefault(x => x.Candidate.IsOpen(iterParams.CurrentDateTime));
if (result == null)
{
throw new UserFriendlyException($"Nie udało się znaleźć żadnego miejsca, gdzie można zjeść w pobliżu o godz: {iterParams.CurrentDateTime}");
}
return(result.Candidate);
}
public async Task <Plan> GenerateAsync(PlanForm planForm)
{
// 1. Create decision array
var DecisionArray = new DecisionArray();
// 2. Create plan object. Validate desitnation and accomodation.
var destinationInfo = await _googlePlaceDetailsApiClient.GetAsync(_googlePlaceDetailsInputFactory.CreateAllUseful(planForm.PlaceId));
var plan = new Plan(destinationInfo.Result.name, destinationInfo.Result.geometry.location.lat, destinationInfo.Result.geometry.location.lng,
(decimal?)destinationInfo.Result.rating, (decimal?)destinationInfo.Result.user_ratings_total, destinationInfo.Result.formatted_address);
if (planForm.HasAccomodationBooked)
{
var accomodationInfo = await _googlePlaceDetailsApiClient.GetAsync(_googlePlaceDetailsInputFactory.CreateAllUseful(planForm.AccomodationId));
plan.PlanAccomodation = new PlanAccomodation(accomodationInfo.Result.geometry.location.lat, accomodationInfo.Result.geometry.location.lng, planForm.AccomodationId,
accomodationInfo.Result.name, accomodationInfo.Result.formatted_address, (decimal?)accomodationInfo.Result.rating, (decimal?)accomodationInfo.Result.user_ratings_total);
var distance = CalculateDistance(plan.Latitude, plan.Longitude, plan.PlanAccomodation.Lat, plan.PlanAccomodation.Lng);
if (distance > MaximumDistanceToAccomodation) //more than 15km
{
throw new UserFriendlyException($"Odległość między celem podróży a miejscem zakwaterowania nie może być większa nić {(int)(MaximumDistanceToAccomodation/1000)} km");
}
}
plan.PlanForm = planForm;
plan.Assumptions = new PlanAssumptions(planForm);
// 3. Generate weight vector based on user preferences
DecisionArray.WeightVector = _weightVectorProvider.Generate(planForm);
plan.PlanFormWeightVector = PlanFormWeightVector.Create(DecisionArray.WeightVector);
// 4. Get plan candidates
var candidates = await _planElementCandidateFactory.GetCandidates(plan, DecisionArray.WeightVector);
//5. Create decision row with values based on candidates
int init = 1;
foreach (var candidate in candidates)
{
DecisionArray.DecisionRows.Add(_decisionRowFactory.Create(candidate, init, plan.StartLocation));
++init;
}
// 6. SCORE FUNCTION -> SAW Normalization (3 types - chosen first) and then calculate Score
var minVector = DecisionArray.GetMinVector();
var maxVector = DecisionArray.GetMaxVector();
foreach (var decisionRow in DecisionArray.DecisionRows)
{
decisionRow.NormalizedScore = _sawMethod.CalculateNormalizedScore(SawNormalizationMethod.LinearFirstType, DecisionArray.WeightVector, decisionRow.DecisionValues, minVector, maxVector);
}
// 7. Clasification
DecisionArray.DecisionRows = DecisionArray.DecisionRows.OrderByDescending(x => x.NormalizedScore).ToList();
int newPos = 1;
foreach (var row in DecisionArray.DecisionRows)
{
row.ScorePosition = newPos;
++newPos;
}
// 8. Create Plan based on decision rows and optimize routes with travel salesman problem
plan.Elements = await _planElementsProvider.GenerateAsync(DecisionArray, plan);
return(plan);
}