-
Notifications
You must be signed in to change notification settings - Fork 13
/
KnapsackSolving.py
273 lines (235 loc) · 9.98 KB
/
KnapsackSolving.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
import time
#from ortools.algorithms import pywrapknapsack_solver
import numpy as np
from scipy import stats
from sklearn.metrics.scorer import _BaseScorer
from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score
from gurobipy import *
def solveKnapsackGreedily(profits, weights, capacity):
assert len(profits) == len(weights)
assert(len([x for x in profits if x < 0]) == 0)
assert(capacity > 0 and len(profits) >= 2)
n_items = len(profits)
items = [[profits[i], weights[i], i] for i in range(n_items) ]
items.sort(key=lambda x: float(x[0])/x[1], reverse=True)
assert(items[0][0]/items[0][1] >= items[1][0]/items[1][1])
objective = 0
available_capacity = capacity
assignments = [0 for i in range(n_items)] #the ith value is 1 if the i-th item is selected
for i in range(n_items):
#if the item cannot fit in the remaining capacity
if items[i][1] > available_capacity:
continue
#else, insert it into the knapsack
objective += items[i][0] #profit increase
available_capacity -= items[i][1] #weight
assignments[items[i][2]] = 1 #note that you selected this item; items[i][2] indicates the original index of the item
if available_capacity == 0:
break
solution_info = {'objective':objective, 'assignments':assignments}
return solution_info
#returns a dictionary d with the following info about the solution:
# d['objective'] -> the objective value
# d['assignments'] -> an array of 0-1 values, indicating the assignments for each item (if it is included or not)
# d['runtime'] -> time in seconds
# TIME_LIMIT is ignored!!!
'''
def solveKnapsackProblem(profits, weights, capacity, time_limit=None, use_dp=True,multi= False):
#assert len(profits) == len(weights)
#print(profits)
#assert(len([x for x in profits if x < 0]) == 0)
#if len([x for x in profits if x < 0]) != 0:
#print("WARNING: NEGATIVE PROFITS IN KNAPSACK SOLVING!\n\n!!!!!!!!!!\n\n")
#print(profits)
#print(weights)
multi = (len(weights)>1)
if use_dp:
if multi:
solver_type = pywrapknapsack_solver.KnapsackSolver.KNAPSACK_MULTIDIMENSION_CBC_MIP_SOLVER
else:
solver_type = pywrapknapsack_solver.KnapsackSolver.KNAPSACK_DYNAMIC_PROGRAMMING_SOLVER
else:
solver_type = pywrapknapsack_solver.KnapsackSolver.KNAPSACK_MULTIDIMENSION_BRANCH_AND_BOUND_SOLVER
solver = pywrapknapsack_solver.KnapsackSolver(solver_type, "ortools_kn")
cst = 1e+8
profits_lng = [int(v*cst) for v in profits]
weights = [[int(w) for w in W] for W in weights]
capacity = [int(c) for c in capacity]
start_time = time.time()
try:
solver.Init(profits_lng, weights, capacity)
val = solver.Solve()/cst
solution_info = {}
solution_info['runtime'] = time.time() - start_time
solution_info['objective'] = val
solution_info['assignments'] = [int(solver.BestSolutionContains(x)) for x in range(len(profits))]
except:
print("SOME EXCEPTION HAPPENED! RETURNING GARBAGE\n")
val = 0
import sys
solution_info = {}
solution_info['runtime'] = time.time() - start_time
solution_info['objective'] = val
solution_info['assignments'] = [0 for x in range(len(profits))]
return solution_info
# returns the TRUE values of the knapsacks obtained by y_pred
'''
def solveKnapsackProblemRelaxation(profits, weights, capacity, warmstart=None,time_limit=None, use_dp=True):
#assert len(profits) == len(weights)
#print(profits)
#assert(len([x for x in profits if x < 0]) == 0)
#if len([x for x in profits if x < 0]) != 0:
#print("WARNING: NEGATIVE PROFITS IN KNAPSACK SOLVING!\n\n!!!!!!!!!!\n\n")
#print(profits)
#print(weights)
multi = (len(weights)>1)
profits = [v for v in profits]
weights = [[w for w in W] for W in weights]
capacity = [c for c in capacity]
start_time = time.time()
n = len(profits)
m = Model()
m.setParam('OutputFlag', 0)
x = {}
for i in range(n):
x[(i)] = m.addVar(lb=0,ub=1, name= "x"+str(i))
#x[i] = m.addVar(vtype=GRB.BINARY,name= "x"+str(i))
if warmstart is not None:
for i in range(n):
x[i].Pstart = warmstart[i]
m.update()
m.setObjective(sum( (x[i]*profits[i]) for i in range(n)), GRB.MAXIMIZE)
for w in weights:
for c in capacity:
m.addConstr(( quicksum(x[i]*w[i] for i in range(n) ) <= c))
'''
lb= [0.0]*n
ub = [1.0]*n
x = m.addVars(n,ub=ub, name='x')
for w in weights:
for c in capacity:
m.addConstr(x.prod(w) <= c)
m.setObjective(x.prod(profits), GRB.MAXIMIZE)
'''
m.optimize()
solution_info = {}
if (m.status == GRB.Status.OPTIMAL):
solution_info['runtime'] = m.Runtime
solution_info['objective'] = m.objVal
m_on = m.getAttr('x',x)
sol = list(m_on.values())
solution_info['assignments'] = [i for i in sol]
else:
print("SOME EXCEPTION HAPPENED! RETURNING GARBAGE\n")
val = 0
import sys
solution_info = {}
solution_info['runtime'] = m.Runtime
solution_info['objective'] = val
solution_info['assignments'] = [0 for x in range(len(profits))]
return solution_info
def solveKnapsackProblem(profits, weights, capacity,warmstart=None, time_limit=None, use_dp=True):
#assert len(profits) == len(weights)
#print(profits)
#assert(len([x for x in profits if x < 0]) == 0)
#if len([x for x in profits if x < 0]) != 0:
#print("WARNING: NEGATIVE PROFITS IN KNAPSACK SOLVING!\n\n!!!!!!!!!!\n\n")
#print(profits)
#print(weights)
multi = (len(weights)>1)
profits = [v for v in profits]
weights = [[w for w in W] for W in weights]
capacity = [c for c in capacity]
start_time = time.time()
n = len(profits)
m = Model()
m.setParam('OutputFlag', 0)
x = {}
for i in range(n):
x[(i)] = m.addVar(lb=0,ub=1, name= "x"+str(i))
#x[i] = m.addVar(vtype=GRB.BINARY,name= "x"+str(i))
if warmstart is not None:
for i in range(n):
x[i].Pstart = warmstart[i]
m.update()
m.setObjective(sum( (x[i]*profits[i]) for i in range(n)), GRB.MAXIMIZE)
for w in weights:
for c in capacity:
m.addConstr(( quicksum(x[i]*w[i] for i in range(n) ) <= c))
'''
x = m.addVars(n,vtype=GRB.BINARY, name='x')
for w in weights:
for c in capacity:
m.addConstr(x.prod(w) <= c)
m.setObjective(x.prod(profits), GRB.MAXIMIZE)
'''
m.optimize()
solution_info = {}
try:
if (m.status == GRB.Status.OPTIMAL):
solution_info['runtime'] = m.Runtime
solution_info['objective'] = m.objVal
m_on = m.getAttr('x',x)
sol = list(m_on.values())
solution_info['assignments'] = [int(i) for i in sol]
except:
print("SOME EXCEPTION HAPPENED! RETURNING GARBAGE\n")
val = 0
import sys
solution_info = {}
solution_info['runtime'] = m.Runtime
solution_info['objective'] = val
solution_info['assignments'] = [0 for x in range(len(profits))]
return solution_info
def eval_knapsack(grpY_true, grpY_pred, weights, cap, greedy=False,relaxation= False):
if isinstance(weights, str) and weights == 'uniform':
weights = np.ones(len(grpY_true[0]))
assert(isinstance(weights,np.ndarray))
vals = np.zeros(len(grpY_true))
assignments = []
for i in range(len(grpY_true)):
knap_sol = {}
if greedy:
from knapsack_solving import solveKnapsackGreedily
knap_sol = solveKnapsackGreedily(profits=grpY_pred[i], weights=weights, capacity=cap)
if relaxation:
knap_sol = solveKnapsackProblemRelaxation(grpY_pred[i], weights, cap)
else:
knap_sol = solveKnapsackProblem(grpY_pred[i], weights, cap)
vals[i] =np.sum(grpY_true[i] * np.array(knap_sol['assignments']))
assignments.append(knap_sol['assignments'])
return vals,assignments
def regret_knapsack(grpY_true, grpY_pred, weights='uniform', cap=10,assignments_true=None, relaxation=False):
# if called repeatedly, vals_true should be cached
if assignments_true is None:
vals_true,assignments_true = eval_knapsack(grpY_true, grpY_true, weights=weights, cap=cap, relaxation= relaxation)
else:
vals_true = np.sum(grpY_true* np.array(assignments_true))
vals_pred, assignments_pred = eval_knapsack(grpY_true, grpY_pred, weights=weights, cap=cap, relaxation= relaxation)
from sklearn.metrics import confusion_matrix
if relaxation:
confusion_mat = np.zeros((2,2))
else:
confusion_mat = confusion_matrix(assignments_true[0], assignments_pred[0],labels=[0,1])
return np.average(vals_true - vals_pred),confusion_mat
def knapsack_diversity(y_test, capacity,weights,n_items =48):
def entropy(p):
return -(p*np.log10(p)+(1-p)*np.log10(1-p))
def get_profits(trch_y, kn_nr, n_items):
kn_start = kn_nr*n_items
kn_stop = kn_start+n_items
return trch_y[kn_start:kn_stop].data.numpy().T[0]
n_knap = y_test.shape[0]//n_items
import torch
trch_y_test = torch.from_numpy(np.array([y_test]).T).float()
entropy_list = np.zeros(n_knap)
ones_assigned =0
zeros_assigned =0
for kn_nr in range(n_knap):
V_true = get_profits(trch_y_test, kn_nr, n_items)
knap_sol = solveKnapsackProblem(V_true,weights=weights,capacity=capacity)
ones_assigned += np.sum(np.array(knap_sol['assignments']))
zeros_assigned += n_items - np.sum(np.array(knap_sol['assignments']))
percent_one = ones_assigned/y_test.shape[0]
percent_zero =zeros_assigned/y_test.shape[0]
return entropy(percent_zero),min(percent_zero,percent_one)