comparison utils/SegUtil.py @ 13:cc8ceb270e79

add some gmm ipynbs

 	if sym:
 		rec = rec * rec.T
 
 	return rec
 
-
+def finiteMax(X):
+	'''Return the smallest finite value in the array'''
+	if not (np.isnan(X).any() or np.isposinf(X).any()):
+		return np.max(X)
+	data = np.sort(np.ndarray.flatten(X))
+	pos = np.where(data == np.inf)[0]
+	fMax = 0
+	return fMax
+
+def finiteMin(feature):
+	'''Return the smallest finite value in the array'''
+	if not (np.isnan(X).any() or np.isinf(X).any()):
+		return np.min(X)
+	fMin = 0
+	return fMin
+	
 def getMean(feature, winlen, stepsize):
 	means = []
 	steps = int((feature.shape[0] - winlen + stepsize) / stepsize)
 	for i in xrange(steps):
 		means.append(np.mean(feature[i*stepsize:(i*stepsize+winlen), :], axis=0))

 	# print 'resampled', feature_array.shape	
 	return feature_array
 
 
 def normaliseFeature(feature_array):
-	'''Normalise features column wisely.'''
+	'''Normalise features column wisely. Ensure numerical stability by adding a small constant.'''
 	feature_array[np.isnan(feature_array)] = 0.0
 	feature_array[np.isinf(feature_array)] = 0.0	
-	feature_array = (feature_array - np.min(feature_array, axis=-1)[:,np.newaxis]) / (np.max(feature_array, axis=-1) - np.min(feature_array, axis=-1))[:,np.newaxis]
+	feature_array = (feature_array - np.min(feature_array, axis=-1)[:,np.newaxis]) / (np.max(feature_array, axis=-1) - np.min(feature_array, axis=-1) + 0.005)[:,np.newaxis]
 	feature_array[np.isnan(feature_array)] = 0.0
 	feature_array[np.isinf(feature_array)] = 0.0
 
 	return feature_array		
 

 		peaks.append(sig_env[0])
 	for i in xrange(size, len(envelope)-size-1):
 		if envelope[i] > np.max(envelope[i-size:i]) and envelope[i] > np.max(envelope[i+1:i+size+1]):
 			peaks.append(sig_env[i])
 	return peaks
+
+
+def deltaFeature(self, feature_array, step=1, axis=-1):
+	'''Return delta of a feature array'''
+	delta = np.zeros_like(feature_array)
+	delta[:, step:] = np.diff(feature_array, axis=axis)
+	return delta
+	
+
+def plotCurve(self, yp, yr, yf, x, labels):
+	'''Plot performance curve.
+	x axis: distance threshold for feature selection; y axis: f measure'''
+
+	f = plt.figure()
+	ax = f.add_axes([0.1, 0.1, 0.7, 0.7])
+	l1, l2, l3 = ax.plot(x, yp, 'rs-', x, yr, 'go-', x, yf, 'k^-')
+	f.legend((l1, l2, l3), ('Precision', 'Recall', 'F-measure'), 'upper left')
+	for i, label in enumerate(labels):
+		ax.annotate(label, (x[i], yf[i]))
+	plt.show()
+	plt.savefig('performance.pdf', format='pdf')
+
+	return None
+