From 48bb47ec86adca90a2194bbae38ce2a6e3ab8d93 Mon Sep 17 00:00:00 2001
From: flatberg <flatberg@pvv.ntnu.no>
Date: Thu, 25 Jan 2007 12:36:32 +0000
Subject: [PATCH] Bugfixed pca

---
 fluents/lib/engines.py | 34 ++++++++++++++++++++++++++++------
 1 file changed, 28 insertions(+), 6 deletions(-)

diff --git a/fluents/lib/engines.py b/fluents/lib/engines.py
index 821dc2b..508027d 100644
--- a/fluents/lib/engines.py
+++ b/fluents/lib/engines.py
@@ -19,14 +19,10 @@ def pca(a, aopt, scale='scores', mode='normal'):
     m, n = a.shape
 
     if m*10.>n:
-        v, s, ut = dot(a.T, a)
-        s = sqrt(s)
-        eigvals = s
-        u = u.T
-        vt = v.T
+        u, s, vt = esvd(a)
     else:
         u, s, vt = svd(a, full_matrices=0)
-        eigvals = (1./m)*s
+    eigvals = (1./m)*s
     T = u*s
     T = T[:,:aopt]
     P = vt[:aopt,:].T
@@ -231,3 +227,29 @@ def bridge(a, b, aopt, scale='scores', mode='normal', r=0):
 
 def m_shape(array):
     return matrix(array).shape
+
+def esvd(data,economy=1):
+    """SVD with the option of economy sized calculation
+    Calculate subspaces of X'X or XX' depending on the shape
+    of the matrix.
+
+    Good for extreme fat or thin matrices
+
+    Numpy supports this by setting full_matrices=0
+    """
+    m, n = data.shape
+    if m>=n:
+        u, s, vt = svd(dot(data.T, data))
+        u = dot(data, vt.T)
+        v = vt.T
+        for i in xrange(n):
+            s[i] = norm(u[:,i])
+            u[:,i] = u[:,i]/s[i]
+    else:
+        u, s, vt = svd(data, data.T)
+        v = dot(u.T, data)
+        for i in xrange(m):
+            s[i] = norm(v[i,:])
+            v[i,:] = v[i,:]/s[i]
+
+    return u, s, v