pulp/pulp_vis.py at master · ajm/pulp · GitHub

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
from sys import exit, stderr, argv
import numpy
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import scipy.sparse
import json
import random
from sklearn import manifold, random_projection

# http://scikit-learn.org/stable/auto_examples/manifold/plot_compare_methods.html
# http://scikit-learn.org/stable/modules/manifold.html

def save_sparse(m, prefix) :
    numpy.save(prefix + '.data.npy',     m.data)
    numpy.save(prefix + '.indices.npy',  m.indices)
    numpy.save(prefix + '.indptr.npy',   m.indptr)
    numpy.save(prefix + '.shape.npy',    m.shape)

def load_data_sparse(prefix) :
    return scipy.sparse.csr_matrix((numpy.load(prefix + '.data.npy'),
                                    numpy.load(prefix + '.indices.npy'),
                                    numpy.load(prefix + '.indptr.npy')),
                                    shape=tuple(numpy.load(prefix + '.shape.npy')))

def load_data() :
    return load_data_sparse('linrel')

def load_json(fname) :
    with open(fname) as f :
        return json.load(f)

def load_features() :
    return load_json('linrel_features.json')

def load_topics() :
    id2topic = load_json('linrel_topics.json')

    #for k in id2topic :
    #    v = id2topic[k]
    #    if "." in v :
    #        id2topic[k] = v.split(".")[0]

    # there are gaps in the arxiv ids, but not in my matrix
    tmp = {}
    keys = [ str(i) for i in sorted([ int(i) for i in id2topic.keys() ]) ]
    for i in range(len(keys)) :
        if id2topic[keys[i]] not in ('cs.SE', 'cs.LG') :
            continue

        tmp[i] = id2topic[keys[i]]

    return tmp

def main() :

    method = argv[1]
    randompicks = int(argv[2])

    methods = ('ltsa','mds','isomap','spectral','tsne')

    if method not in methods :
        print >> stderr, "%s not recognised (%s)" % (method, ', '.join(methods))
        exit(1)

    # load + constants
    print >> stderr, "loading data..."
    data = load_data()
    num_articles, num_features = data.shape
    num_neighbours = 10
    id2topic = load_topics()

    # random selection of articles
    if randompicks != -1 :
        articles = sorted(random.sample(id2topic.keys(), randompicks))
    else :
        articles = id2topic.keys()

    data = data[ numpy.array(articles) ]
    num_articles, num_features = data.shape

    id2topic = dict([ (i, id2topic[i]) for i in articles ])

    #num_components = len(set(id2topic.values()))
    num_components = 2
    colours = [ id2topic[k] for k in sorted(id2topic.keys()) ]

    print >> stderr, "loaded %d articles x %d features" % (num_articles, num_features)
    print >> stderr, "  ... with %d components" % num_components

    #print >> stderr, colours[:10]

    # transform
    print >> stderr, "transforming..."

    if method == 'ltsa' :
        t = manifold.LocallyLinearEmbedding(num_neighbours,
                                            num_components,
                                            eigen_solver='auto',
                                            method='ltsa')

    elif method == 'mds' :
        t = manifold.MDS(num_components,
                         max_iter=100,
                         n_init=1)

    elif method == 'isomap' :
        t = manifold.Isomap(num_neighbours,
                            num_components)

    elif method == 'spectral' :
        t = manifold.SpectralEmbedding(n_components=num_components,
                                       n_neighbors=num_neighbours)

    elif method == 'tsne' :
        t = manifold.TSNE(n_components=num_components,
                          init='pca',
                          random_state=0)

    X = t.fit_transform(data.toarray())

#    save_sparse(X, method)
#    return 0


#    X = random_projection.SparseRandomProjection(n_components=2, random_state=42).fit_transform(data)
    print >> stderr, "transform done, shape =", X.shape

    # normalise
    x_min, x_max = numpy.min(X, 0), numpy.max(X, 0)
    X = (X - x_min) / (x_max - x_min)

    #X = X.toarray()

#    print >> stderr, colours.shape

#    print >> stderr, X.shape
#    print >> stderr, X[:,0].shape
#    print >> stderr, X[:,1].shape

    df = pd.DataFrame(dict(x=X[:,0], y=X[:,1], colours=colours))

    sns.lmplot('x', 'y', data=df, hue='colours', fit_reg=False)

    # plot
#    plt.scatter(X[:, 0],
#                X[:, 1],
#                c=colours,
#                cmap=plt.cm.Spectral)

#    plt.legend()
    plt.show()

if __name__ == '__main__' :
    try :
        exit(main())
    except KeyboardInterrupt :
        print >> stderr, "Killed by User...\n"
        exit(1)