Mercurial Hosting > traffic-intelligence

comparison scripts/learn-poi.py @ 805:180b6b0231c0

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
added saving/loading points of interests
author Nicolas Saunier <nicolas.saunier@polymtl.ca>
date Thu, 09 Jun 2016 15:36:21 -0400
parents 0a428b449b80
children 181bcb6dad3a
comparison
equal deleted inserted replaced
801:c5f98916297e 805:180b6b0231c0
9 import storage, ml 9 import storage, ml
10 10
11 parser = argparse.ArgumentParser(description='The program learns and displays Gaussians fit to beginnings and ends of object trajectories (based on Mohamed Gomaa Mohamed 2015 PhD). TODO: save the data') 11 parser = argparse.ArgumentParser(description='The program learns and displays Gaussians fit to beginnings and ends of object trajectories (based on Mohamed Gomaa Mohamed 2015 PhD). TODO: save the data')
12 parser.add_argument('-d', dest = 'databaseFilename', help = 'name of the Sqlite database file', required = True) 12 parser.add_argument('-d', dest = 'databaseFilename', help = 'name of the Sqlite database file', required = True)
13 parser.add_argument('-t', dest = 'trajectoryType', help = 'type of trajectories to display', choices = ['feature', 'object'], default = 'object') 13 parser.add_argument('-t', dest = 'trajectoryType', help = 'type of trajectories to display', choices = ['feature', 'object'], default = 'object')
14 parser.add_argument('-n', dest = 'nClusters', help = 'number of point clusters', required = True, type = int) 14 parser.add_argument('-norigins', dest = 'nOriginClusters', help = 'number of clusters for trajectory origins', required = True, type = int)
15 parser.add_argument('-ndestinations', dest = 'nDestinationClusters', help = 'number of clusters for trajectory destinations (=norigins if not provided)', type = int)
15 parser.add_argument('--covariance-type', dest = 'covarianceType', help = 'type of covariance of Gaussian model', default = "full") 16 parser.add_argument('--covariance-type', dest = 'covarianceType', help = 'type of covariance of Gaussian model', default = "full")
16 parser.add_argument('-w', dest = 'worldImageFilename', help = 'filename of the world image') 17 parser.add_argument('-w', dest = 'worldImageFilename', help = 'filename of the world image')
17 parser.add_argument('-u', dest = 'pixelsPerUnit', help = 'number pixels per unit of distance', type = float, default = 1.) 18 parser.add_argument('-u', dest = 'unitsPerPixel', help = 'number of units of distance per pixel', type = float, default = 1.)
18 19
19 args = parser.parse_args() 20 args = parser.parse_args()
20 21
21 objects = storage.loadTrajectoriesFromSqlite(args.databaseFilename, args.trajectoryType) 22 objects = storage.loadTrajectoriesFromSqlite(args.databaseFilename, args.trajectoryType)
22 23
27 ends.append(o.getPositionAt(int(o.length())-1).aslist()) 28 ends.append(o.getPositionAt(int(o.length())-1).aslist())
28 29
29 beginnings = np.array(beginnings) 30 beginnings = np.array(beginnings)
30 ends = np.array(ends) 31 ends = np.array(ends)
31 32
32 gmm = mixture.GMM(n_components=args.nClusters, covariance_type = args.covarianceType) 33 nDestinationClusters = args.nDestinationClusters
33 beginningModel=gmm.fit(beginnings) 34 if args.nDestinationClusters is None:
34 gmm = mixture.GMM(n_components=args.nClusters, covariance_type = args.covarianceType) 35 nDestinationClusters = args.nOriginClusters
35 endModel=gmm.fit(ends)
36 36
37 fig = plt.figure() 37 gmmId=0
38 if args.worldImageFilename is not None and args.pixelsPerUnit is not None: 38 for nClusters, points, gmmType in zip([args.nOriginClusters, nDestinationClusters],
39 img = plt.imread(args.worldImageFilename) 39 [beginnings, ends],
40 plt.imshow(img) 40 ['beginning', 'end']):
41 ml.plotGMMClusters(beginningModel, beginnings, fig, nPixelsPerUnit = args.pixelsPerUnit) 41 # estimation
42 plt.axis('equal') 42 gmm = mixture.GMM(n_components=nClusters, covariance_type = args.covarianceType)
43 plt.title('Origins') 43 model=gmm.fit(beginnings)
44 print('Origin Clusters:\n{}'.format(ml.computeClusterSizes(beginningModel.predict(beginnings), range(args.nClusters)))) 44 if not model.converged_:
45 45 print('Warning: model for '+gmmType+' points did not converge')
46 fig = plt.figure() 46 # plot
47 if args.worldImageFilename is not None and args.pixelsPerUnit is not None: 47 fig = plt.figure()
48 img = plt.imread(args.worldImageFilename) 48 if args.worldImageFilename is not None and args.unitsPerPixel is not None:
49 plt.imshow(img) 49 img = plt.imread(args.worldImageFilename)
50 ml.plotGMMClusters(endModel, ends, fig, nPixelsPerUnit = args.pixelsPerUnit) 50 plt.imshow(img)
51 plt.axis('equal') 51 labels = ml.plotGMMClusters(model, points, fig, nUnitsPerPixel = args.unitsPerPixel)
52 plt.title('Destinations') 52 plt.axis('image')
53 print('Destination Clusters:\n{}'.format(ml.computeClusterSizes(endModel.predict(ends), range(args.nClusters)))) 53 plt.title(gmmType)
54 print(gmmType+' Clusters:\n{}'.format(ml.computeClusterSizes(labels, range(model.n_components))))
55 # save
56 storage.savePOIs(args.databaseFilename, model, gmmType, gmmId)
57 gmmId += 1
58
59 # fig = plt.figure()
60 # if args.worldImageFilename is not None and args.pixelsPerUnit is not None:
61 # img = plt.imread(args.worldImageFilename)
62 # plt.imshow(img)
63 # ml.plotGMMClusters(, , fig, nPixelsPerUnit = args.pixelsPerUnit)
64 # plt.axis('equal')
65 # plt.title()
66 # print('Destination Clusters:\n{}'.format(ml.computeClusterSizes(endModel.predict(ends), range(args.nClusters))))