def build_meal(dishes, lower_boundaries, upper_boundaries):
#reject if there simply aren't enough meals
if len(dishes) < 10:
return "NOT ENOUGH DISHES TO MAKE A MEAL"
nutrient_reject = {x: 0 for x in TARGET_NUTRIENTS}
myMeal = Meal(lower_boundaries)
#descending list of foods with most of each respective nutrient
sorted_nutrients = {
nutrient: sorted(dishes,
key=lambda x: x.nutrition[nutrient],
reverse=True)
for nutrient in TARGET_NUTRIENTS
}
sorted_indices = {nutrient: 0 for nutrient in TARGET_NUTRIENTS}
bestMeal = 0
bestScore = math.inf
for i in range(100):
highest_unmet = myMeal.nutrient_totals()[0]
if sorted_indices[highest_unmet] >= len(
sorted_nutrients[highest_unmet]):
dish = random.choice(sorted_nutrients[highest_unmet])
else:
dish = sorted_nutrients[highest_unmet][
sorted_indices[highest_unmet]]
sorted_indices[highest_unmet] += 1
while myMeal.containsDish(dish):
if sorted_indices[highest_unmet] >= len(
sorted_nutrients[highest_unmet]):
dish = random.choice(sorted_nutrients[highest_unmet])
else:
dish = sorted_nutrients[highest_unmet][
sorted_indices[highest_unmet]]
sorted_indices[highest_unmet] += 1
# print("We need {} so we add {}".format(highest_unmet,dish.name))
# print(myMeal)
myMeal.addDish(dish)
# print(myMeal)
constraint = myMeal.under_upper_bounds(upper_boundaries)
if constraint[0]:
if myMeal.all_nutrients_met():
return myMeal, True
else:
# print("Too much of the {} nutrient".format(constraint[1]))
myMeal.removeDish(dish.name)
# print("Removed {}".format(dish.name))
nutrient_reject[constraint[1]] += 1
if nutrient_reject[constraint[1]] > 3 and not myMeal.isEmpty():
sorted_meal = myMeal.sort_by_nutrient(constraint[1])
myMeal.removeDish(sorted_meal[-1])
# print("Also removed {} because of excess {}".format(sorted_meal[-1],
# constraint[1]))
nutrient_reject[constraint[1]] = 0
if myMeal.meal_score() < bestScore:
bestScore = myMeal.meal_score()
bestMeal = Meal(myMeal.boundaries.copy())
bestMeal.items = myMeal.items.copy()
bestMeal.totals = myMeal.totals.copy()
# print(myMeal.meal_score())
# print()
return bestMeal, False
