From 88c2b29d765b75e3f5dce962c1aff1621ff26963 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?C=C3=A9dric=20Traizet?= <cedric.traizet@c-s.fr>
Date: Wed, 4 Sep 2019 17:17:33 +0200
Subject: [PATCH] ENH: add parameter to disable whitening in pca
---
.../app/otbDimensionalityReduction.cxx | 5 +++++
.../include/otbPCAImageFilter.h | 4 ++++
.../include/otbPCAImageFilter.hxx | 14 +++++++++-----
3 files changed, 18 insertions(+), 5 deletions(-)
diff --git a/Modules/Applications/AppDimensionalityReduction/app/otbDimensionalityReduction.cxx b/Modules/Applications/AppDimensionalityReduction/app/otbDimensionalityReduction.cxx
index 24f5951f6b..3e5cbf8e7a 100644
--- a/Modules/Applications/AppDimensionalityReduction/app/otbDimensionalityReduction.cxx
+++ b/Modules/Applications/AppDimensionalityReduction/app/otbDimensionalityReduction.cxx
@@ -130,6 +130,10 @@ private:
AddParameter(ParameterType_OutputFilename, "method.pca.outeigenvalues", "Output file containing eigenvalues (txt format)");
SetParameterDescription("method.pca.outeigenvalues", "Output file containing eigenvalues (txt format).");
MandatoryOff("method.pca.outeigenvalues");
+ AddParameter(ParameterType_Bool, "method.pca.whiten", "Perform pca whitening");
+ SetParameterDescription("method.pca.whiten", "Perform whitening and ensure uncorrelated outputs with unit component wise variances");
+ SetParameterInt("method.pca.whiten", 1);
+ MandatoryOff("method.pca.whiten");
AddChoice("method.napca", "NA-PCA");
SetParameterDescription("method.napca", "Noise Adjusted Principal Component Analysis.");
@@ -257,6 +261,7 @@ private:
filter->SetInput(GetParameterFloatVectorImage("in"));
filter->SetNumberOfPrincipalComponentsRequired(nbComp);
+ filter->SetWhitening(GetParameterInt("method.pca.whiten"));
// Center AND reduce the input data.
if (normalize)
diff --git a/Modules/Filtering/DimensionalityReduction/include/otbPCAImageFilter.h b/Modules/Filtering/DimensionalityReduction/include/otbPCAImageFilter.h
index 3368163846..b90faae536 100644
--- a/Modules/Filtering/DimensionalityReduction/include/otbPCAImageFilter.h
+++ b/Modules/Filtering/DimensionalityReduction/include/otbPCAImageFilter.h
@@ -97,6 +97,9 @@ public:
*/
itkSetMacro(NumberOfPrincipalComponentsRequired, unsigned int);
itkGetMacro(NumberOfPrincipalComponentsRequired, unsigned int);
+
+ itkSetMacro(Whitening, bool);
+ itkGetMacro(Whitening, bool);
itkGetMacro(CovarianceEstimator, CovarianceEstimatorFilterType *);
itkGetMacro(Transformer, TransformFilterType *);
@@ -197,6 +200,7 @@ protected:
bool m_GivenCovarianceMatrix;
bool m_GivenTransformationMatrix;
bool m_IsTransformationMatrixForward;
+ bool m_Whitening;
VectorType m_MeanValues;
VectorType m_StdDevValues;
diff --git a/Modules/Filtering/DimensionalityReduction/include/otbPCAImageFilter.hxx b/Modules/Filtering/DimensionalityReduction/include/otbPCAImageFilter.hxx
index 3c2487119c..51633d1eaf 100644
--- a/Modules/Filtering/DimensionalityReduction/include/otbPCAImageFilter.hxx
+++ b/Modules/Filtering/DimensionalityReduction/include/otbPCAImageFilter.hxx
@@ -39,7 +39,7 @@ PCAImageFilter< TInputImage, TOutputImage, TDirectionOfTransformation >
this->SetNumberOfRequiredInputs(1);
m_NumberOfPrincipalComponentsRequired = 0;
-
+ m_Whitening = true;
m_UseNormalization = false;
m_UseVarianceForNormalization = false;
m_GivenMeanValues = false;
@@ -181,7 +181,7 @@ PCAImageFilter< TInputImage, TOutputImage, TDirectionOfTransformation >
{
for (unsigned int c = 0; c < numberOfComponent; ++c)
{
- m_CovarianceMatrix(r,c) = cov(r,c) / (std::sqrt(cov(r,r)) *std::sqrt( cov(c,c)));
+ m_CovarianceMatrix(r,c) = cov(r,c) / std::sqrt(cov(r,r)*cov(c,c));
}
}
}
@@ -376,19 +376,23 @@ PCAImageFilter< TInputImage, TOutputImage, TDirectionOfTransformation >
m_EigenValues.SetSize( m_NumberOfPrincipalComponentsRequired );
for ( unsigned int i = 0; i < m_NumberOfPrincipalComponentsRequired; ++i )
- m_EigenValues[m_NumberOfPrincipalComponentsRequired - 1 - i] = static_cast< RealType >( valP(i, i) );
+ m_EigenValues[m_NumberOfPrincipalComponentsRequired - 1 - i] = static_cast< RealType >( vectValP[i] );
/* We used normalized PCA */
for ( unsigned int i = 0; i < valP.rows(); ++i )
{
if ( valP(i, i) > 0. )
{
- valP(i, i) = 1. / std::sqrt( valP(i, i) );
+ if (m_Whitening)
+ valP(i, i) = 1. / std::sqrt( valP(i, i) );
}
else if ( valP(i, i) < 0. )
{
otbMsgDebugMacro( << "ValP(" << i << ") neg : " << valP(i, i) << " taking abs value" );
- valP(i, i) = 1. / std::sqrt( std::abs( valP(i, i) ) );
+ if (m_Whitening)
+ valP(i, i) = 1. / std::sqrt( std::abs( valP(i, i) ) );
+ else
+ valP(i, i) = std::abs( valP(i, i));
}
else
{
--
GitLab