ell = []
for i in range(len(theta)):
pos = theta[i].mean
vals,vecs = _eigsorted(theta[i].var)
th = np.degrees(np.arctan2(*vecs[:,0][::-1]))
for v in vol:
width,height = 2.0*np.sqrt(chi2.ppf(v,2))*np.sqrt(vals)
ell.append(Ellipse(xy=pos,width=width,height=height,angle=th))
for i,e in enumerate(ell):
ax.add_artist(e)
e.set_facecolor(my_color_map(i))
e.set_alpha(0.5)
ax.set_xlim(xmin, xmax)
ax.set_ylim(ymin, ymax)
plt.show()
if __name__ == '__main__':
from generate_data import generate_timeseries_set as gtss
nStates = 4
dim = 2
NumTS = 5
theta, truestates, data = gtss(nStates,dim,NumTS,minl=50, maxl=100)
generate_Fplot(nStates,truestates)
generate_timeseriesplot(data,truestates)
generate_emissions(theta,data)
from plotter import *
from parameters import *
# ------------------------------------------------- CREATE TOY DATA!
if __name__ == '__main__':
print 'Creating some toy data...\n'
nStates = 4
numTS = 5
dim = 2
# First, we'll create some toy data
# 5 sequences, each of length between 50 and 500.
# Each sequences selects from 4 behaviors,
# and switches among its selected set over time.
# We'll use num=4 behaviors, each of which defines a distinct Gaussian
# emission distribution (with dim=2 dimensions).
emissions, trueStates, data = gtss(nStates, dim, numTS, minl=50, maxl=500)
# Visualize the raw data time series
# with background colored by "true" hidden state
'''generate_timeseriesplot(data,trueStates)
# Visualize the "true" generating parameters
# Feat matrix F (binary numTS x nStates matrix )
generate_Fplot(nStates,trueStates)
# Emission parameters theta (Gaussian 2D contours)
generate_emissions(emissions,data)
# ------------------------------------------------- RUN MCMC INFERENCE!
modelP = ModelParams_BPHMM(data)
modelP.bpM.gamma = 2
from plotter import * from parameters import * # ------------------------------------------------- CREATE TOY DATA! if __name__ == '__main__': print 'Creating some toy data...\n' nStates = 4 numTS = 5 dim = 2 # First, we'll create some toy data # 5 sequences, each of length between 50 and 500. # Each sequences selects from 4 behaviors, # and switches among its selected set over time. # We'll use num=4 behaviors, each of which defines a distinct Gaussian # emission distribution (with dim=2 dimensions). emissions, trueStates, data = gtss(nStates, dim, numTS, minl=50, maxl=500) # Visualize the raw data time series # with background colored by "true" hidden state '''generate_timeseriesplot(data,trueStates) # Visualize the "true" generating parameters # Feat matrix F (binary numTS x nStates matrix ) generate_Fplot(nStates,trueStates) # Emission parameters theta (Gaussian 2D contours) generate_emissions(emissions,data) # ------------------------------------------------- RUN MCMC INFERENCE! modelP = ModelParams_BPHMM(data) modelP.bpM.gamma = 2
ell = []
for i in range(len(theta)):
pos = theta[i].mean
vals, vecs = _eigsorted(theta[i].var)
th = np.degrees(np.arctan2(*vecs[:, 0][::-1]))
for v in vol:
width, height = 2.0 * np.sqrt(chi2.ppf(v, 2)) * np.sqrt(vals)
ell.append(Ellipse(xy=pos, width=width, height=height, angle=th))
for i, e in enumerate(ell):
ax.add_artist(e)
e.set_facecolor(my_color_map(i))
e.set_alpha(0.5)
ax.set_xlim(xmin, xmax)
ax.set_ylim(ymin, ymax)
plt.show()
if __name__ == '__main__':
from generate_data import generate_timeseries_set as gtss
nStates = 4
dim = 2
NumTS = 5
theta, truestates, data = gtss(nStates, dim, NumTS, minl=50, maxl=100)
generate_Fplot(nStates, truestates)
generate_timeseriesplot(data, truestates)
generate_emissions(theta, data)