From d954829a40748b108653c96b8283fb5432ce001a Mon Sep 17 00:00:00 2001
From: Max Zwiessele <ibinbei@gmail.com>
Date: Thu, 9 Oct 2014 08:40:15 +0100
Subject: [PATCH] [pca] missing data is now handled as mean

---
 GPy/util/pca.py | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)

diff --git a/GPy/util/pca.py b/GPy/util/pca.py
index 046f47d7..996e6f48 100644
--- a/GPy/util/pca.py
+++ b/GPy/util/pca.py
@@ -11,14 +11,16 @@ try:
 except:
     pass
 from numpy.linalg.linalg import LinAlgError
+from operator import setitem
+import itertools
 
 class pca(object):
     """
     pca module with automatic primal/dual determination.
     """
     def __init__(self, X):
-        self.mu = X.mean(0)
-        self.sigma = X.std(0)
+        self.mu = None
+        self.sigma = None
 
         X = self.center(X)
 
@@ -39,6 +41,13 @@ class pca(object):
         """
         Center `X` in pca space.
         """
+        X = X.copy()
+        inan = numpy.isnan(X)
+        if self.mu is None:
+            X_ = numpy.ma.masked_array(X, inan)
+            self.mu = X_.mean(0).base
+            self.sigma = X_.std(0).base
+        reduce(lambda y,x: setitem(x[0], x[1], x[2]), itertools.izip(X.T, inan.T, self.mu), None)
         X = X - self.mu
         X = X / numpy.where(self.sigma == 0, 1e-30, self.sigma)
         return X
@@ -94,7 +103,7 @@ class pca(object):
                 fignum=None, cmap=None, # @UndefinedVariable
                 ** kwargs):
         """
-        Plot dimensions `dimensions` with given labels against each other in 
+        Plot dimensions `dimensions` with given labels against each other in
         PC space. Labels can be any sequence of labels of dimensions X.shape[0].
         Labels can be drawn with a subsequent call to legend()
         """