/
utils.py
80 lines (73 loc) · 2.73 KB
/
utils.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
import cPickle as pickle
import datasets, config
import numpy as np
import pandas as pd
from sklearn.externals import joblib
def load_from_cache(filename):
filename = '%s/%s.pkl' % (config.CACHEDIR, filename)
if config.CACHETYPE == 'joblib':
obj = joblib.load(filename)
elif config.CACHETYPE == 'pickle':
f = open(filename)
obj = pickle.load(f)
f.close()
else:
raise ValueError('Unkown CACHETYPE %s, use only pickle or joblib' % config.CACHETYPE)
return obj
def save_to_cache(obj, filename):
filename = '%s/%s.pkl' % (config.CACHEDIR, filename)
if config.CACHETYPE == 'joblib':
joblib.dump(obj, filename, compress=9)
elif config.CACHETYPE == 'pickle':
f = open('cache/%s.pkl' % filename, 'w')
pickle.dump(obj, f, 2)
f.close()
else:
raise ValueError('Unkown CACHETYPE %s, use only pickle or joblib' % config.CACHETYPE)
def create_submission(filename, pred, ids=None):
data = ['id,num_views,num_votes,num_comments']
pred = pred.astype('S100')
if ids is None:
ids = datasets.load_dataset('TestIDS')
for id, p in zip(ids, pred):
data.append('%i,' % (id) + ','.join(p))
data = '\n'.join(data)
f = open('%s/%s' %(config.SUBMITDIR, filename), 'w')
f.write(data)
f.close()
def make_vw(data, targets, filename):
"""
Helper method to create a vowpal wabbit dataset from data and targets and
save it to filename
"""
s = []
for yi, xi in zip(targets, data):
xis = ' '.join(['f%i:%f' % (f, xi[0,f]) for f in xi.nonzero()[1]])
s.append('%f | %s' %(yi, xis))
f = open(filename, 'w')
f.write('\n'.join(s))
f.close()
def greedy_feature_selection(model, features, j):
selected_features = set()
score_hist = []
ycv = exp(y_cv) - 1
while len(selected_features) < len(features):
scores = []
for i in range(len(features)):
if i not in selected_features:
feats = list(selected_features) + [i]
if len(feats) == 1:
ttfs = features[i]
else:
ttfs = data_transforms.drop_disjoint(sparse.hstack((
features[feats])).tocsr(), targets)
X_train_pre = ttfs[:n_train]
X_train = X_train_pre[:int(n_train*0.8)]
X_cv = X_train_pre[int(n_train*0.8):]
model.fit(X_train[-keep:], y_train[-keep:])
cv = exp(ridge.predict(X_cv)) - 1
scores.append((rmsle(postprocess_pred(cv)[:,j], ycv[:,j]), feats, i))
print scores[-1]
selected_features.add(min(scores)[2])
score_hist.append(min(scores))
return score_hist