From d5df0b5fc40470cc75731572f98a8338381f8d3b Mon Sep 17 00:00:00 2001
From: Petr Baudis <pasky@ucw.cz>
Date: Tue, 2 Mar 2010 09:38:00 +0100
Subject: [PATCH] PCA: Print also projection information on the output

That is, mapping from PCA vectors to input vector patterns
---
 gostyle.py | 15 +++++++++++++--
 pca.py     | 13 ++++++++++---
 2 files changed, 23 insertions(+), 5 deletions(-)

diff --git a/gostyle.py b/gostyle.py
index c21f2d5..55b102f 100755
--- a/gostyle.py
+++ b/gostyle.py
@@ -88,6 +88,8 @@ try:
 			return list(self.pca(array([vector]))[0])
 		def process_list_of_vectors(self, list_of_vectors):
 			return [ list(vec) for vec in self.pca(array(list_of_vectors)) ]
+		def get_projection_info(self):
+			return self.pca.get_recmatrix()
 
 except ImportError, e:
 	print >>sys.stderr, "Warning: %s. PCA will not work."%(str(e))
@@ -127,12 +129,15 @@ class OccurenceVectorGenerator(VectorGenerator):
 	def generate_top_pattern_dict(self):
 		rexp=re.compile(Const.pat_file_regexp)
 		self.top_pattern_dict = {}
+		self.top_pattern_str = {}
 		i = 0
 		input_file = open(self.filename)
 		for line in input_file:
 			if i >= self.output_dim:
 				break
-			self.top_pattern_dict[rexp.match(line).group(2)] = i
+			s = rexp.match(line).group(2)
+			self.top_pattern_dict[s] = i
+			self.top_pattern_str[i] = s
 			i += 1
 		input_file.close()
 
@@ -161,6 +166,9 @@ class OccurenceVectorGenerator(VectorGenerator):
 			raise RuntimeError
 		return vector
 
+	def stringof(self, i):
+		return self.top_pattern_str[i]
+
 class Rescale(VectorToVector):
 	"""Class that rescales vectors to a given interval!"""
 	def __init__(self, a=-1.0, b=1.0):
@@ -183,9 +191,12 @@ class InputVectorGenerator(VectorGenerator):
 	one is mapped to 1.0 and the rest is mapped relatively on the scale (1.0,-1.0). See `__call__' function.
 	"""
 	def __init__(self, *args, **kwargs):
-		self.gen = Compose(OccurenceVectorGenerator(*args, **kwargs), Rescale(-1.0, 1.0))
+		self.ovg = OccurenceVectorGenerator(*args, **kwargs)
+		self.gen = Compose(self.ovg, Rescale(-1.0, 1.0))
 	def __call__(self, *args, **kwargs):
 		return self.gen(*args, **kwargs)
+	def ovg(self):
+		return self.ovg
 
 def linear_combination(list_of_vectors, coefs):
 	if len(list_of_vectors) != len(coefs):
diff --git a/pca.py b/pca.py
index 16105ad..189c582 100755
--- a/pca.py
+++ b/pca.py
@@ -25,13 +25,13 @@ if __name__ == '__main__':
 
 	### Objects creating input and output vectors when called
 	print >>sys.stderr, "Creating input vector generator from main pat file:", main_pat_filename
-	i = InputVectorGenerator(main_pat_filename, num_features)
+	ivg = InputVectorGenerator(main_pat_filename, num_features)
 
 	# Create pairs of (input vector, player name)
 	input_vectors = []
 	for name in players: 
-		#input_vectors += [i( Data.pat_files_folder + name)]
-		input_vectors += [[float(occ) for occ in i(Data.pat_files_folder + name)]]
+		#input_vectors += [ivg( Data.pat_files_folder + name)]
+		input_vectors += [[float(occ) for occ in ivg(Data.pat_files_folder + name)]]
 
 	if len(input_vectors) == 0:
 		print >>sys.stderr, "No input vectors.", main_pat_filename
@@ -66,6 +66,13 @@ if __name__ == '__main__':
 		#print name_to_print,
 		#print_vector(vector)
 
+	print "\nProjection info:"
+	P = pca.get_projection_info()
+	for y in xrange(1, P.shape[0]):
+		for x in xrange(1, P.shape[1]):
+			print y, x, P[y,x], ivg.ovg.stringof(x)
+	#print P
+
 	print >> sys.stderr, "\nNow print that by:"
 	print >> sys.stderr, 'gnuplot> set xrange[1:%d]'%(pca.pca.output_dim+1)
 	print >> sys.stderr, 'gnuplot> plot "./pca.data" using 2:3:1 with labels font "arial,10" left point pt 4 offset 1,0'
-- 
2.11.4.GIT