def __init__(self, x, kernel=None):
x = np.asarray(x)
if x.ndim == 1:
x = x[:, None]
nobs, n_series = x.shape
# print "%s dimensions" % n_series
if kernel is None:
kernel = kernels.Gaussian() # no meaningful bandwidth yet
if n_series > 1:
if isinstance(kernel, kernels.CustomKernel):
kernel = kernels.NdKernel(n_series, kernels=kernel)
self.kernel = kernel
self.n = n_series # TODO change attribute
self.x = x
def __init__(self, kernel=kernel.Gaussian(), params={}):
"""
Initialization of a new instance of KRLS.
"""
self.kernel = kernel # Instance of a kernel to use
# initialize with some reasonable values
self.adopt_thresh = 0.02 # approximate linear dependence threshold [0,1]
self.dico_max_size = 15 # maximum size of the dictionary
self.adaptive = True # flag if elimination is data-adaptive
self.alpha = None
self.p = [[1]]
self.gamma = 0.01 # intrinsic regularization (ridge term coefficient)
self.dico = None # the dictionary of relevant states
self.target = None # the associated targets
self.k = None # the kernel matrix
self.kinv = None # the inverse kernel matrix
self.forget_rate = 0.0 # how fast we want to forget what we did see
self.dico_idx = [] # index when we added what
# Allow overriding the attributes through the params dictionary
for k in params:
if hasattr(self, k):
setattr(self, k, params[k])
if __name__ == "__main__":
PLOT = True
from numpy import random
import matplotlib.pyplot as plt
import bandwidths as bw
# 1 D case
random.seed(142)
x = random.standard_t(4.2, size=50)
h = bw.bw_silverman(x)
#NOTE: try to do it with convolution
support = np.linspace(-10, 10, 512)
kern = kernels.Gaussian(h=h)
kde = KDE(x, kern)
print kde.density(1.015469)
print 0.2034675
Xs = np.arange(-10, 10, 0.1)
if PLOT:
fig = plt.figure()
ax = fig.add_subplot(111)
ax.plot(Xs, kde(Xs), "-")
ax.set_ylim(-10, 10)
ax.set_ylim(0, 0.4)
plt.show()
# 2 D case
# from statsmodels.sandbox.nonparametric.testdata import kdetest
def __init__(self, x, y, Kernel=None):
if Kernel is None:
Kernel = kernel.Gaussian()
self.Kernel = Kernel
self.x = np.array(x)
self.y = np.array(y)