-
Notifications
You must be signed in to change notification settings - Fork 1
/
utils.py
63 lines (51 loc) · 1.81 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
#! /usr/bin/env python
# -*- coding: utf-8 -*-
import numpy as np
import sktensor as skt
def rmse(Y1,Y2):
return np.sqrt(np.linalg.norm(Y1-Y2)**2/Y2.size)
class Kernel(object):
def __init__(self,name,param=None, normalize=False):
assert name in ['rbf','linear','poly'], 'kernel not implemented'
if name == 'rbf':
self.f = lambda x,y: np.exp(-1*param * np.linalg.norm(x-y)**2)
self.param = param
self.name = name
if name == 'linear':
self.f = np.dot
self.param = param
self.name = name
if name == 'poly':
self.f = lambda x,y: (param[1] + x.T.dot(y))**param[0]
self.param = param
self.name = name
self.normalize = normalize
def __call__(self,x,y):
if self.normalize:
z = np.sqrt(self.f(x,x)*self.f(y,y))
else:
z = 1
return self.f(x,y) / z
def __str__(self):
return self.name + ' (param=%s, norm=%s)' % (str(self.param), str(self.normalize))
def gram_matrix(self,X,X_test=None):
if isinstance(X,skt.dtensor):
X = X.unfold(0)
if X_test is None:
n = X.shape[0]
K = np.zeros([n,n])
for i in range(n):
for j in range(i,n):
tmp = self(X[i,:],X[j,:])
K[i,j] = tmp
K[j,i] = tmp
else:
if isinstance(X_test,skt.dtensor):
X_test = X_test.unfold(0)
K = np.zeros([X_test.shape[0], X.shape[0]])
for i in range(X_test.shape[0]):
for j in range(X.shape[0]):
K[i,j] = self(X_test[i,:],X[j,:])
return K
def rbf_kernel(gamma):
return lambda x,y: np.exp(-1*gamma * np.linalg.norm(x-y)**2)