anydev/anyway
anydev/anyway
0
0
Fork 0
Code Issues Pull Requests 3 Actions Packages Projects Releases Wiki Activity

fixed the optimizer_v4

Browse Source
This commit is contained in:
Helldragon67 2024-07-07 16:24:15 +02:00
parent e71c92da40
commit d4e964c5d4
7 changed files with 358 additions and 113 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
backend/src/amenities/nature.am
View File

@ -8,4 +8,5 @@ geological
'tourism'='alpine_hut'
'tourism'='viewpoint'
'tourism'='zoo'
#'tourism'='artwork'
'waterway'='waterfall'

1
backend/src/amenities/sightseeing.am
View File

@ -1,6 +1,7 @@
'tourism'='museum'
'tourism'='attraction'
'tourism'='gallery'
'tourism'='artwork'
historic
'amenity'='planetarium'
'amenity'='place_of_worship'

21
backend/src/optimizer_v4.py
View File

@ -179,6 +179,10 @@ def init_ub_dist(landmarks: List[Landmark], max_steps: int):
for j, spot2 in enumerate(landmarks) :
t = get_time(spot1.location, spot2.location, detour, speed)
dist_table[j] = t
closest = sorted(dist_table)[:22]
for i, dist in enumerate(dist_table) :
if dist not in closest :
dist_table[i] = 10000000
A_ub += dist_table
c = c*len(landmarks)
@ -186,7 +190,7 @@ def init_ub_dist(landmarks: List[Landmark], max_steps: int):
# Constraint to respect only one travel per landmark. Also caps the total number of visited landmarks
def respect_number(L:int, A_ub, b_ub):
def respect_number(L: int, A_ub, b_ub, max_landmarks):
ones = [1]*L
zeros = [0]*L
@ -195,10 +199,11 @@ def respect_number(L:int, A_ub, b_ub):
A_ub = np.vstack((A_ub, h))
b_ub.append(1)
# Read the parameters from the file
with open (os.path.dirname(os.path.abspath(__file__)) + '/parameters/optimizer.params', "r") as f :
parameters = json.loads(f.read())
max_landmarks = parameters['max landmarks']
if max_landmarks is None :
# Read the parameters from the file
with open (os.path.dirname(os.path.abspath(__file__)) + '/parameters/optimizer.params', "r") as f :
parameters = json.loads(f.read())
max_landmarks = parameters['max landmarks']
A_ub = np.vstack((A_ub, ones*L))
b_ub.append(max_landmarks+1)
@ -361,13 +366,13 @@ def link_list_simple(ordered_visit: List[Landmark])-> List[Landmark] :
# Main optimization pipeline
def solve_optimization (landmarks :List[Landmark], max_steps: int, printing_details: bool) :
def solve_optimization (landmarks :List[Landmark], max_steps: int, printing_details: bool, max_landmarks = None) :
L = len(landmarks)
# SET CONSTRAINTS FOR INEQUALITY
c, A_ub, b_ub = init_ub_dist(landmarks, max_steps) # Add the distances from each landmark to the other
A_ub, b_ub = respect_number(L, A_ub, b_ub) # Respect max number of visits (no more possible stops than landmarks).
A_ub, b_ub = respect_number(L, A_ub, b_ub, max_landmarks) # Respect max number of visits (no more possible stops than landmarks).
A_ub, b_ub = break_sym(L, A_ub, b_ub) # break the 'zig-zag' symmetry
# SET CONSTRAINTS FOR EQUALITY
@ -392,7 +397,7 @@ def solve_optimization (landmarks :List[Landmark], max_steps: int, printing_deta
i = 0
timeout = 80
while len(circle) != 0 and i < timeout:
#A_ub, b_ub = prevent_config(res.x, A_ub, b_ub)
A_ub, b_ub = prevent_config(res.x, A_ub, b_ub)
A_ub, b_ub = break_cricle(order, len(landmarks), A_ub, b_ub)
res = linprog(c, A_ub=A_ub, b_ub=b_ub, A_eq=A_eq, b_eq = b_eq, bounds=x_bounds, method='highs', integrality=3)
order, circle = is_connected(res.x)

4
backend/src/parameters/landmarks_manager.params
View File

@ -1,8 +1,8 @@
{
"city bbox side" : 3,
"city bbox side" : 10,
"radius close to" : 50,
"church coeff" : 0.9,
"park coeff" : 1.2,
"tag coeff" : 10,
"N important" : 30
"N important" : 40
}

121
backend/src/refiner.py
View File

@ -1,20 +1,15 @@
from collections import defaultdict
from heapq import heappop, heappush
from itertools import permutations
import os, json
from shapely import buffer, LineString, Point, Polygon, MultiPoint, convex_hull, concave_hull, LinearRing
from shapely import buffer, LineString, Point, Polygon, MultiPoint, concave_hull
from typing import List, Tuple
from scipy.spatial import KDTree
from math import pi
import networkx as nx
from structs.landmarks import Landmark
from landmarks_manager import take_most_important
from optimizer_v4 import solve_optimization, link_list_simple, print_res, get_time
from optimizer_v2 import generate_path, generate_path2
# Create corridor from tour
def create_corridor(landmarks: List[Landmark], width: float) :
corrected_width = (180*width)/(6371000*pi)
@ -25,6 +20,7 @@ def create_corridor(landmarks: List[Landmark], width: float) :
return obj
# Create linestring from tour
def create_linestring(landmarks: List[Landmark])->List[Point] :
points = []
@ -35,11 +31,13 @@ def create_linestring(landmarks: List[Landmark])->List[Point] :
return LineString(points)
# Check if some coordinates are in area. Used for the corridor
def is_in_area(area: Polygon, coordinates) -> bool :
point = Point(coordinates)
return point.within(area)
# Function to determine if two landmarks are close to each other
def is_close_to(location1: Tuple[float], location2: Tuple[float]):
"""Determine if two locations are close by rounding their coordinates to 3 decimals."""
absx = abs(location1[0] - location2[0])
@ -49,6 +47,7 @@ def is_close_to(location1: Tuple[float], location2: Tuple[float]):
#return (round(location1[0], 3), round(location1[1], 3)) == (round(location2[0], 3), round(location2[1], 3))
# Rearrange some landmarks in the order of visit
def rearrange(landmarks: List[Landmark]) -> List[Landmark]:
i = 1
@ -65,6 +64,8 @@ def rearrange(landmarks: List[Landmark]) -> List[Landmark]:
return landmarks
# Simple nearest neighbour planner to try to fix the path
def find_shortest_path_through_all_landmarks(landmarks: List[Landmark]) -> Tuple[List[Landmark], Polygon]:
# Read from data
@ -106,6 +107,8 @@ def find_shortest_path_through_all_landmarks(landmarks: List[Landmark]) -> Tuple
return path, path_poly
# Returns a list of minor landmarks around the planned path to enhance experience
def get_minor_landmarks(all_landmarks: List[Landmark], visited_landmarks: List[Landmark], width: float) -> List[Landmark] :
second_order_landmarks = []
@ -122,22 +125,7 @@ def get_minor_landmarks(all_landmarks: List[Landmark], visited_landmarks: List[L
return take_most_important(second_order_landmarks, len(visited_landmarks))
"""def refine_optimization(landmarks: List[Landmark], base_tour: List[Landmark], max_time: int, print_infos: bool) -> List[Landmark] :
minor_landmarks = get_minor_landmarks(landmarks, base_tour, 200)
if print_infos : print("There are " + str(len(minor_landmarks)) + " minor landmarks around the predicted path")
full_set = base_tour[:-1] + minor_landmarks # create full set of possible landmarks (without finish)
full_set.append(base_tour[-1]) # add finish back
new_tour = solve_optimization(full_set, max_time, print_infos)
return new_tour"""
# Try fix the shortest path using shapely
def fix_using_polygon(tour: List[Landmark])-> List[Landmark] :
coords = []
@ -150,12 +138,18 @@ def fix_using_polygon(tour: List[Landmark])-> List[Landmark] :
tour_poly = Polygon(coords)
better_tour_poly = tour_poly.buffer(0)
xs, ys = better_tour_poly.exterior.xy
try :
xs, ys = better_tour_poly.exterior.xy
if len(xs) != len(tour) :
if len(xs) != len(tour) :
better_tour_poly = concave_hull(MultiPoint(coords)) # Create concave hull with "core" of tour leaving out start and finish
xs, ys = better_tour_poly.exterior.xy
except :
better_tour_poly = concave_hull(MultiPoint(coords)) # Create concave hull with "core" of tour leaving out start and finish
xs, ys = better_tour_poly.exterior.xy
# reverse the xs and ys
xs.reverse()
ys.reverse()
@ -183,6 +177,7 @@ def fix_using_polygon(tour: List[Landmark])-> List[Landmark] :
return better_tour
# Second stage of the optimization. Use linear programming again to refine the path
def refine_optimization(landmarks: List[Landmark], base_tour: List[Landmark], max_time: int, print_infos: bool) -> List[Landmark] :
# Read from the file
@ -199,7 +194,7 @@ def refine_optimization(landmarks: List[Landmark], base_tour: List[Landmark], ma
full_set.append(base_tour[-1]) # add finish back
# get a new tour
new_tour = solve_optimization(full_set, max_time, False)
new_tour = solve_optimization(full_set, max_time, False, max_landmarks)
new_tour, new_dist = link_list_simple(new_tour)
better_tour, better_poly = find_shortest_path_through_all_landmarks(new_tour)
@ -218,78 +213,14 @@ def refine_optimization(landmarks: List[Landmark], base_tour: List[Landmark], ma
if print_infos :
print("\n\n\nRefined tour (result of second stage optimization): ")
print_res(final_tour, len(full_set))
total_score = 0
for elem in final_tour :
total_score += elem.attractiveness
print("\nTotal score : " + str(total_score))
return final_tour
def refine_path(landmarks: List[Landmark], base_tour: List[Landmark], max_time: int, print_infos: bool) -> List[Landmark] :
print("\nRefining the base tour...")
# Read from the file
with open (os.path.dirname(os.path.abspath(__file__)) + '/parameters/optimizer.params', "r") as f :
parameters = json.loads(f.read())
max_landmarks = parameters['max landmarks'] + 3
"""if len(base_tour)-2 >= max_landmarks :
return base_tour"""
minor_landmarks = get_minor_landmarks(landmarks, base_tour, 200)
if print_infos : print("Using " + str(len(minor_landmarks)) + " minor landmarks around the predicted path")
full_set = base_tour + minor_landmarks # create full set of possible landmarks
print("\nRefined tour (result of second stage optimization): ")
new_path = generate_path2(full_set, max_time, max_landmarks)
return new_path
# If a tour is not connected
def correct_path(tour: List[Landmark]) -> List[Landmark] :
# Read the parameters from the file
with open (os.path.dirname(os.path.abspath(__file__)) + '/parameters/optimizer.params', "r") as f :
parameters = json.loads(f.read())
detour = parameters['detour factor']
speed = parameters['average walking speed']
G = nx.Graph()
coords = []
landmap = {}
for i, landmark in enumerate(tour) :
coords.append(landmark.location)
landmap[i] = landmark
G.add_node(i, pos=landmark.location, weight=landmark.attractiveness)
kdtree = KDTree(coords)
k = 3
for node, coord in coords:
indices = kdtree.query(coord, k + 1)[1] # k+1 because the closest neighbor is the node itself
for idx in indices[1:]: # skip the first one (itself)
neighbor = list(coords)[idx]
distance = get_time(coord, coords[neighbor], detour, speed)
G.add_edge(node, neighbor, weight=distance)
path = nx.approximation.traveling_salesman_problem(G, weight='weight', cycle=True)
if len(path) != len(tour) :
print("nope")
lis = [landmap[id] for id in path]
lis, tot_dist = link_list_simple(lis)
print_res(lis, len(tour))
return path

306
backend/src/refiner_v2.py Normal file
View File

@ -0,0 +1,306 @@
from collections import defaultdict
from heapq import heappop, heappush
from itertools import permutations
import os, json
from shapely import buffer, LineString, Point, Polygon, MultiPoint, convex_hull, concave_hull, LinearRing
from typing import List, Tuple
from scipy.spatial import KDTree
from math import pi
import networkx as nx
from structs.landmarks import Landmark
from landmarks_manager import take_most_important
from optimizer_v4 import solve_optimization, link_list_simple, print_res, get_time
from optimizer_v2 import generate_path, generate_path2
def create_corridor(landmarks: List[Landmark], width: float) :
corrected_width = (180*width)/(6371000*pi)
path = create_linestring(landmarks)
obj = buffer(path, corrected_width, join_style="mitre", cap_style="square", mitre_limit=2)
return obj
def create_linestring(landmarks: List[Landmark])->List[Point] :
points = []
for landmark in landmarks :
points.append(Point(landmark.location))
return LineString(points)
def is_in_area(area: Polygon, coordinates) -> bool :
point = Point(coordinates)
return point.within(area)
def is_close_to(location1: Tuple[float], location2: Tuple[float]):
"""Determine if two locations are close by rounding their coordinates to 3 decimals."""
absx = abs(location1[0] - location2[0])
absy = abs(location1[1] - location2[1])
return absx < 0.001 and absy < 0.001
#return (round(location1[0], 3), round(location1[1], 3)) == (round(location2[0], 3), round(location2[1], 3))
def rearrange(landmarks: List[Landmark]) -> List[Landmark]:
i = 1
while i < len(landmarks):
j = i+1
while j < len(landmarks):
if is_close_to(landmarks[i].location, landmarks[j].location) and landmarks[i].name not in ['start', 'finish'] and landmarks[j].name not in ['start', 'finish']:
# If they are not adjacent, move the j-th element to be adjacent to the i-th element
if j != i + 1:
landmarks.insert(i + 1, landmarks.pop(j))
break # Move to the next i-th element after rearrangement
j += 1
i += 1
return landmarks
def find_shortest_path_through_all_landmarks(landmarks: List[Landmark]) -> Tuple[List[Landmark], Polygon]:
# Read from data
with open (os.path.dirname(os.path.abspath(__file__)) + '/parameters/optimizer.params', "r") as f :
parameters = json.loads(f.read())
detour = parameters['detour factor']
speed = parameters['average walking speed']
# Step 1: Find 'start' and 'finish' landmarks
start_idx = next(i for i, lm in enumerate(landmarks) if lm.name == 'start')
finish_idx = next(i for i, lm in enumerate(landmarks) if lm.name == 'finish')
start_landmark = landmarks[start_idx]
finish_landmark = landmarks[finish_idx]
# Step 2: Create a list of unvisited landmarks excluding 'start' and 'finish'
unvisited_landmarks = [lm for i, lm in enumerate(landmarks) if i not in [start_idx, finish_idx]]
# Step 3: Initialize the path with the 'start' landmark
path = [start_landmark]
coordinates = [landmarks[start_idx].location]
current_landmark = start_landmark
# Step 4: Use nearest neighbor heuristic to visit all landmarks
while unvisited_landmarks:
nearest_landmark = min(unvisited_landmarks, key=lambda lm: get_time(current_landmark.location, lm.location, detour, speed))
path.append(nearest_landmark)
coordinates.append(nearest_landmark.location)
current_landmark = nearest_landmark
unvisited_landmarks.remove(nearest_landmark)
# Step 5: Finally add the 'finish' landmark to the path
path.append(finish_landmark)
coordinates.append(landmarks[finish_idx].location)
path_poly = Polygon(coordinates)
return path, path_poly
def get_minor_landmarks(all_landmarks: List[Landmark], visited_landmarks: List[Landmark], width: float) -> List[Landmark] :
second_order_landmarks = []
visited_names = []
area = create_corridor(visited_landmarks, width)
for visited in visited_landmarks :
visited_names.append(visited.name)
for landmark in all_landmarks :
if is_in_area(area, landmark.location) and landmark.name not in visited_names:
second_order_landmarks.append(landmark)
return take_most_important(second_order_landmarks, len(visited_landmarks))
"""def refine_optimization(landmarks: List[Landmark], base_tour: List[Landmark], max_time: int, print_infos: bool) -> List[Landmark] :
minor_landmarks = get_minor_landmarks(landmarks, base_tour, 200)
if print_infos : print("There are " + str(len(minor_landmarks)) + " minor landmarks around the predicted path")
full_set = base_tour[:-1] + minor_landmarks # create full set of possible landmarks (without finish)
full_set.append(base_tour[-1]) # add finish back
new_tour = solve_optimization(full_set, max_time, print_infos)
return new_tour"""
def fix_using_polygon(tour: List[Landmark])-> List[Landmark] :
coords = []
coords_dict = {}
for landmark in tour :
coords.append(landmark.location)
if landmark.name != 'finish' :
coords_dict[landmark.location] = landmark
tour_poly = Polygon(coords)
better_tour_poly = tour_poly.buffer(0)
try :
xs, ys = better_tour_poly.exterior.xy
if len(xs) != len(tour) :
better_tour_poly = concave_hull(MultiPoint(coords)) # Create concave hull with "core" of tour leaving out start and finish
xs, ys = better_tour_poly.exterior.xy
except :
better_tour_poly = concave_hull(MultiPoint(coords)) # Create concave hull with "core" of tour leaving out start and finish
xs, ys = better_tour_poly.exterior.xy
# reverse the xs and ys
xs.reverse()
ys.reverse()
better_tour = [] # List of ordered visit
name_index = {} # Maps the name of a landmark to its index in the concave polygon
# Loop through the polygon and generate the better (ordered) tour
for i,x in enumerate(xs[:-1]) :
y = ys[i]
better_tour.append(coords_dict[tuple((x,y))])
name_index[coords_dict[tuple((x,y))].name] = i
# Scroll the list to have start in front again
start_index = name_index['start']
better_tour = better_tour[start_index:] + better_tour[:start_index]
# Append the finish back and correct the time to reach
better_tour.append(tour[-1])
# Rearrange only if polygon
better_tour = rearrange(better_tour)
return better_tour
def refine_optimization(landmarks: List[Landmark], base_tour: List[Landmark], max_time: int, print_infos: bool) -> List[Landmark] :
# Read from the file
with open (os.path.dirname(os.path.abspath(__file__)) + '/parameters/optimizer.params', "r") as f :
parameters = json.loads(f.read())
max_landmarks = parameters['max landmarks'] + 4
minor_landmarks = get_minor_landmarks(landmarks, base_tour, 200)
if print_infos : print("Using " + str(len(minor_landmarks)) + " minor landmarks around the predicted path")
# full set of visitable landmarks
full_set = base_tour[:-1] + minor_landmarks # create full set of possible landmarks (without finish)
full_set.append(base_tour[-1]) # add finish back
# get a new tour
new_tour = solve_optimization(full_set, max_time, False, max_landmarks)
new_tour, new_dist = link_list_simple(new_tour)
better_tour, better_poly = find_shortest_path_through_all_landmarks(new_tour)
if base_tour[0].location == base_tour[-1].location and not better_poly.is_valid :
better_tour = fix_using_polygon(better_tour)
better_tour, better_dist = link_list_simple(better_tour)
if new_dist < better_dist :
final_tour = new_tour
else :
final_tour = better_tour
if print_infos :
print("\n\n\nRefined tour (result of second stage optimization): ")
print_res(final_tour, len(full_set))
total_score = 0
for elem in final_tour :
total_score += elem.attractiveness
print("\nTotal score : " + str(total_score))
return final_tour
def refine_path(landmarks: List[Landmark], base_tour: List[Landmark], max_time: int, print_infos: bool) -> List[Landmark] :
print("\nRefining the base tour...")
# Read from the file
with open (os.path.dirname(os.path.abspath(__file__)) + '/parameters/optimizer.params', "r") as f :
parameters = json.loads(f.read())
max_landmarks = parameters['max landmarks'] + 3
"""if len(base_tour)-2 >= max_landmarks :
return base_tour"""
minor_landmarks = get_minor_landmarks(landmarks, base_tour, 200)
if print_infos : print("Using " + str(len(minor_landmarks)) + " minor landmarks around the predicted path")
full_set = base_tour + minor_landmarks # create full set of possible landmarks
print("\nRefined tour (result of second stage optimization): ")
new_path = generate_path2(full_set, max_time, max_landmarks)
return new_path
# If a tour is not connected
def correct_path(tour: List[Landmark]) -> List[Landmark] :
# Read the parameters from the file
with open (os.path.dirname(os.path.abspath(__file__)) + '/parameters/optimizer.params', "r") as f :
parameters = json.loads(f.read())
detour = parameters['detour factor']
speed = parameters['average walking speed']
G = nx.Graph()
coords = []
landmap = {}
for i, landmark in enumerate(tour) :
coords.append(landmark.location)
landmap[i] = landmark
G.add_node(i, pos=landmark.location, weight=landmark.attractiveness)
kdtree = KDTree(coords)
k = 3
for node, coord in coords:
indices = kdtree.query(coord, k + 1)[1] # k+1 because the closest neighbor is the node itself
for idx in indices[1:]: # skip the first one (itself)
neighbor = list(coords)[idx]
distance = get_time(coord, coords[neighbor], detour, speed)
G.add_edge(node, neighbor, weight=distance)
path = nx.approximation.traveling_salesman_problem(G, weight='weight', cycle=True)
if len(path) != len(tour) :
print("nope")
lis = [landmap[id] for id in path]
lis, tot_dist = link_list_simple(lis)
print_res(lis, len(tour))
return path

17
backend/src/tester.py
View File

@ -73,7 +73,7 @@ def test4(coordinates: tuple[float, float]) -> List[Landmark]:
sightseeing=Preference(
name='sightseeing',
type=LandmarkType(landmark_type='sightseeing'),
score = 5),
score = 0),
nature=Preference(
name='nature',
type=LandmarkType(landmark_type='nature'),
@ -105,21 +105,22 @@ def test4(coordinates: tuple[float, float]) -> List[Landmark]:
with open (os.path.dirname(os.path.abspath(__file__)) + '/parameters/optimizer.params', "r") as f :
parameters = json.loads(f.read())
max_landmarks = parameters['max landmarks']
max_walking_time = 120 # minutes
detour = 30 # minutes
max_walking_time = 120 # minutes
detour = 10 # minutes
# First stage optimization
#base_tour = solve_optimization(landmarks_short, max_walking_time, True)
base_tour = solve_optimization(landmarks_short, max_walking_time, True)
#base_tour = solve_optimization(landmarks_short, max_walking_time, True)
# First stage using NetworkX
base_tour = generate_path(landmarks_short, max_walking_time, max_landmarks)
#base_tour = generate_path(landmarks_short, max_walking_time, max_landmarks)
# Second stage using linear optimization
refined_tour = refine_optimization(landmarks, base_tour, max_walking_time+detour, True)
if detour != 0 :
refined_tour = refine_optimization(landmarks, base_tour, max_walking_time+detour, True)
# Second stage using NetworkX
#refined_tour = refine_path(landmarks, base_tour, max_walking_time+detour, True)
@ -130,8 +131,8 @@ def test4(coordinates: tuple[float, float]) -> List[Landmark]:
return refined_tour
#test4(tuple((48.8344400, 2.3220540))) # Café Chez César
test4(tuple((48.8344400, 2.3220540))) # Café Chez César
#test4(tuple((48.8375946, 2.2949904))) # Point random
#test4(tuple((47.377859, 8.540585))) # Zurich HB
test4(tuple((45.7576485, 4.8330241))) # Lyon Bellecour
#test4(tuple((45.7576485, 4.8330241))) # Lyon Bellecour
#test3('Vienna, Austria')