def __init__(self, estimates, S): self.estimates = estimates self.Delta = np.vstack( [ self._H(s) for s in S ] ) #self.varphi = svm( data=self.Delta, Lambda=.00005, gamma=[.125,.25,.5,1,2,4,8,16] ) self.varphi = svm( data=self.Delta, Lambda=.0000005, gamma=[16,32,64,128,256,512] )
def single():
fig = plt.figure()
xrange = (-3.0, 9.0)
yrange = (-3.0, 9.0)
xstep = 0.1
ystep = 0.1
samples1 = np.vstack(
[
np.random.multivariate_normal(mean=np.array([0, 0]), cov=np.array(np.identity(2)), size=np.array([50])),
np.random.multivariate_normal(mean=np.array([0, 6]), cov=np.array(np.identity(2)), size=np.array([50])),
]
)
phi_1 = svm(samples1, Lambda=0.05, gamma=[0.125, 0.25, 0.5, 1, 2, 4, 8, 16])
a = fig.add_subplot(2, 2, 1)
phi_1.contourPlot(fig=a, xrange=xrange, yrange=yrange, xstep=xstep, ystep=ystep)
samples2 = np.vstack(
[
np.random.multivariate_normal(mean=np.array([0, 0]), cov=np.array(np.identity(2)), size=np.array([50])),
np.random.multivariate_normal(mean=np.array([6, 0]), cov=np.array(np.identity(2)), size=np.array([50])),
]
)
phi_2 = svm(samples2, Lambda=0.05, gamma=[0.125, 0.25, 0.5, 1, 2, 4, 8, 16])
b = fig.add_subplot(2, 2, 2)
phi_2.contourPlot(fig=b, xrange=xrange, yrange=yrange, xstep=xstep, ystep=ystep)
samples = list()
for i in range(10):
samples.append(
np.vstack(
[
np.random.multivariate_normal(
mean=np.array([0, 0]), cov=np.array(np.identity(2)), size=np.array([50])
),
np.random.multivariate_normal(
mean=np.array([6, 0]), cov=np.array(np.identity(2)), size=np.array([50])
),
]
)
)
samples.append(
np.vstack(
[
np.random.multivariate_normal(
mean=np.array([0, 0]), cov=np.array(np.identity(2)), size=np.array([50])
),
np.random.multivariate_normal(
mean=np.array([0, 6]), cov=np.array(np.identity(2)), size=np.array([50])
),
]
)
)
for i in range(10):
samples.append(
np.vstack(
[
np.random.multivariate_normal(
mean=np.array([0, 0]), cov=np.array(np.identity(2)), size=np.array([5])
),
np.random.multivariate_normal(
mean=np.array([6, 0]), cov=np.array(np.identity(2)), size=np.array([5])
),
]
)
)
samples.append(
np.vstack(
[
np.random.multivariate_normal(
mean=np.array([0, 0]), cov=np.array(np.identity(2)), size=np.array([5])
),
np.random.multivariate_normal(
mean=np.array([0, 6]), cov=np.array(np.identity(2)), size=np.array([5])
),
]
)
)
e = engine(estimates=(phi_1, phi_2), S=samples)
test1 = np.vstack(
[np.random.multivariate_normal(mean=np.array([0, 0]), cov=np.array(np.identity(2)), size=np.array([10]))]
)
c = fig.add_subplot(2, 2, 3)
e.contourPlot(
S=test1, fig=c, xrange=xrange, yrange=yrange, xstep=xstep, ystep=ystep, title="derived distribution 1"
)
# e.varphi.contourPlot( c, xrange=(-.1,0.), yrange=(-.1,0.), xstep=.005,ystep=.005 )
# e.varphiPlot(c)
# c.hist(e.varphi.pdf( e.Delta ) )
test2 = np.vstack(
[np.random.multivariate_normal(mean=np.array([0, 6]), cov=np.array(np.identity(2)), size=np.array([10]))]
)
c = fig.add_subplot(2, 2, 4)
e.contourPlot(
S=test2, fig=c, xrange=xrange, yrange=yrange, xstep=xstep, ystep=ystep, title="derived distribution 2"
)
plt.show()
