def LA_od_costs(factors, output, verbose=0): ''' compute the OD costs for UE, SO, and UE-K where the cognitive cost is K=3000 and with different demand: alpha * demand for demand in factors save OD costs into csv array with columns demand, X1_so, X1_ue_k, X1_ue, X2_so, X2_ue_k, X2_ue, ... ''' net, demand, node, geom = load_LA_3() demand[:, 2] = demand[:, 2] / 4000. fs_so = np.loadtxt('data/LA/so_single_class.csv', delimiter=',', skiprows=1) fs_ue_k = np.loadtxt('data/LA/ue_k_single_class.csv', delimiter=',', skiprows=1) fs_ue = np.loadtxt('data/LA/ue_single_class.csv', delimiter=',', skiprows=1) costs = [] for i in range(len(factors)): costs.append(cost(fs_so[:, i], net)) costs.append(cost(fs_ue_k[:, i], net)) costs.append(cost(fs_ue[:, i], net)) free_flow_OD_costs(net, costs, demand, output, verbose)
def compute_metrics_beta(alpha, beta, f, net, d, feat, subset, out, row, fs=None, net2=None, \ length_unit='Mile', time_unit='Minute'): ''' Save in the numpy array 'out' at the specific 'row' the following metrics - average cost for non-routed - average cost for routed - average cost - average cost on a subset (e.g. local routes) - average cost outside of a subset (e.g. non-local routes) - total gas emissions - total gas emissions on a subset (e.g. local routes) - total gas emissions outside of a subset (e.g. non-local routes) - total flow in the network - total flow in the network on a subset (e.g. local routes) - total flow in the network outside of a subset (e.g. non-local routes) ''' if length_unit == 'Meter': lengths = feat[:, 1] / 1609.34 # convert into miles elif length_unit == 'Mile': lengths = feat[:, 1] if time_unit == 'Minute': a = 60.0 elif time_unit == 'Second': a = 3600. b = 60. / a speed = a * np.divide(lengths, np.maximum(cost(f, net), 10e-8)) co2 = np.multiply(gas_emission(speed), lengths) out[row, 0] = alpha out[row, 1] = beta out[row, 4] = b * average_cost(f, net, d) out[row, 5] = b * average_cost_subset(f, net, d, subset) out[row, 6] = out[row, 3] - out[row, 4] out[row, 7] = co2.dot(f) / f.dot(lengths) out[row, 8] = np.multiply(co2, subset).dot(f) / f.dot(lengths) out[row, 9] = out[row, 6] - out[row, 7] out[row, 10] = np.sum(np.multiply(f, lengths)) * 4000. out[row, 11] = np.sum(np.multiply(np.multiply(f, lengths), subset)) * 4000. out[row, 12] = out[row, 9] - out[row, 10] if alpha == 0.0: out[row, 2] = b * average_cost(f, net, d) out[row, 3] = b * average_cost_all_or_nothing(f, net, d) return if alpha == 1.0: L = all_or_nothing_assignment(cost(f, net2), net, d) out[row, 2] = b * cost(f, net).dot(L) / np.sum(d[:, 2]) out[row, 3] = b * average_cost(f, net, d) return out[row, 2] = b * cost(f, net).dot(fs[:, 0]) / np.sum( (1 - alpha) * d[:, 2]) out[row, 3] = b * cost(f, net).dot(fs[:, 1]) / np.sum(alpha * d[:, 2])
def compute_metrics_beta(alpha, beta, f, net, d, feat, subset, out, row, fs=None, net2=None, \ length_unit='Mile', time_unit='Minute'): ''' Save in the numpy array 'out' at the specific 'row' the following metrics - average cost for non-routed - average cost for routed - average cost - average cost on a subset (e.g. local routes) - average cost outside of a subset (e.g. non-local routes) - total gas emissions - total gas emissions on a subset (e.g. local routes) - total gas emissions outside of a subset (e.g. non-local routes) - total flow in the network - total flow in the network on a subset (e.g. local routes) - total flow in the network outside of a subset (e.g. non-local routes) ''' if length_unit == 'Meter': lengths = feat[:,1] / 1609.34 # convert into miles elif length_unit == 'Mile': lengths = feat[:,1] if time_unit == 'Minute': a = 60.0 elif time_unit == 'Second': a = 3600. b = 60./a speed = a * np.divide(lengths, np.maximum(cost(f, net), 10e-8)) co2 = np.multiply(gas_emission(speed), lengths) out[row,0] = alpha out[row,1] = beta out[row,4] = b * average_cost(f, net, d) out[row,5] = b * average_cost_subset(f, net, d, subset) out[row,6] = out[row,3] - out[row,4] out[row,7] = co2.dot(f) / f.dot(lengths) out[row,8] = np.multiply(co2, subset).dot(f) / f.dot(lengths) out[row,9] = out[row,6] - out[row,7] out[row,10] = np.sum(np.multiply(f, lengths)) * 4000. out[row,11] = np.sum(np.multiply(np.multiply(f, lengths), subset)) * 4000. out[row,12] = out[row,9] - out[row,10] if alpha == 0.0: out[row,2] = b * average_cost(f, net, d) out[row,3] = b * average_cost_all_or_nothing(f, net, d) return if alpha == 1.0: L = all_or_nothing_assignment(cost(f, net2), net, d) out[row,2] = b * cost(f, net).dot(L) / np.sum(d[:,2]) out[row,3] = b * average_cost(f, net, d) return out[row,2] = b * cost(f, net).dot(fs[:,0]) / np.sum((1-alpha)*d[:,2]) out[row,3] = b * cost(f, net).dot(fs[:,1]) / np.sum(alpha*d[:,2])
def frank_wolfe_on_I210(): ''' Frank-Wolfe on I210 ''' graph = np.loadtxt('data/I210_attack_net.csv', delimiter=',', skiprows=1) demand = np.loadtxt('data/I210_od.csv', delimiter=',', skiprows=1) demand[:, 2] = 1. * demand[:, 2] / 4000 # run solver f = solver_3(graph, demand, max_iter=1000, q=50, display=1, stop=1e-2) # display cost for a, b in zip(cost(f, graph), f * 4000): print a, b # visualization node = np.loadtxt('data/I210_node.csv', delimiter=',', skiprows=1) # extract features: 'capacity', 'length', 'fftt' feat = extract_features('data/I210_attack_Sketch_net.csv') ratio = cost_ratio(f, graph) # merge features with the cost ratios features = np.zeros((feat.shape[0], 4)) features[:, :3] = feat features[:, 3] = ratio # join features with (lat1,lon1,lat2,lon2) links = process_links(graph, node, features) color = features[:, 3] # we choose the costs names = ['capacity', 'length', 'fftt', 'tt_over_fftt'] geojson_link(links, names, color)
def frank_wolfe_on_I210(): ''' Frank-Wolfe on I210 ''' graph = np.loadtxt('data/I210_attack_net.csv', delimiter=',', skiprows=1) demand = np.loadtxt('data/I210_od.csv', delimiter=',', skiprows=1) demand[:,2] = 1. * demand[:,2] / 4000 # run solver f = solver_3(graph, demand, max_iter=1000, q=50, display=1, stop=1e-2) # display cost for a,b in zip(cost(f, graph), f*4000): print a,b # visualization node = np.loadtxt('data/I210_node.csv', delimiter=',', skiprows=1) # extract features: 'capacity', 'length', 'fftt' feat = extract_features('data/I210_attack_Sketch_net.csv') ratio = cost_ratio(f, graph) # merge features with the cost ratios features = np.zeros((feat.shape[0],4)) features[:,:3] = feat features[:,3] = ratio # join features with (lat1,lon1,lat2,lon2) links = process_links(graph, node, features) color = features[:,3] # we choose the costs names = ['capacity', 'length', 'fftt', 'tt_over_fftt'] geojson_link(links, names, color)
def LA_od_costs(factors, output, verbose=0): ''' compute the OD costs for UE, SO, and UE-K where the cognitive cost is K=3000 and with different demand: alpha * demand for demand in factors save OD costs into csv array with columns demand, X1_so, X1_ue_k, X1_ue, X2_so, X2_ue_k, X2_ue, ... ''' net, demand, node, geom = load_LA_3() demand[:,2] = demand[:,2] / 4000. fs_so = np.loadtxt('data/LA/so_single_class.csv', delimiter=',', skiprows=1) fs_ue_k = np.loadtxt('data/LA/ue_k_single_class.csv', delimiter=',', skiprows=1) fs_ue = np.loadtxt('data/LA/ue_single_class.csv', delimiter=',', skiprows=1) costs = [] for i in range(len(factors)): costs.append(cost(fs_so[:,i],net)) costs.append(cost(fs_ue_k[:,i],net)) costs.append(cost(fs_ue[:,i],net)) free_flow_OD_costs(net, costs, demand, output, verbose)
def check_LA_result(): net, demand, node, features = load_LA_2() demand[:, 2] = demand[:, 2] / 4000. f = np.loadtxt('data/LA/LA_output_4.csv', delimiter=',', skiprows=0) costs = cost(f, net) cr = cost_ratio(f, net) print np.sort(cr)[-20:] for row in range(net.shape[0]): if cr[row] >= 10.: print cr[row] print net[row, :3], features[row, :] L = all_or_nothing_assignment(costs, net, demand) print costs.dot(L) / np.sum(demand[:, 2])
def check_LA_result(): net, demand, node, features = load_LA_2() demand[:,2] = demand[:,2] / 4000. f = np.loadtxt('data/LA/LA_output_4.csv', delimiter=',', skiprows=0) costs = cost(f, net) cr = cost_ratio(f, net) print np.sort(cr)[-20:] for row in range(net.shape[0]): if cr[row] >= 10.: print cr[row] print net[row,:3], features[row,:] L = all_or_nothing_assignment(costs, net, demand) print costs.dot(L) / np.sum(demand[:,2])
def test_cost(self): net = np.loadtxt('data/braess_net.csv', delimiter=',', skiprows=1) net[:,5] = np.array([2.]*5) flow = np.array([0., 1., 2., 3., 4.]) result = np.array([0., 3., 8., 19., 36.]) self.assertTrue(np.linalg.norm(result - cost(flow, net)) < 1e-8)
def test_cost(self): net = np.loadtxt('data/braess_net.csv', delimiter=',', skiprows=1) net[:, 5] = np.array([2.] * 5) flow = np.array([0., 1., 2., 3., 4.]) result = np.array([0., 3., 8., 19., 36.]) self.assertTrue(np.linalg.norm(result - cost(flow, net)) < 1e-8)