EMMAforDoubleMatrix

net.maizegenetics.stats.EMMA.EMMAforDoubleMatrix

```
public class EMMAforDoubleMatrix
```
Class for implementation of the EMMA algorithm often used in GWAS. Initially developer by Kang et al Genetics March 2008 178:1709-1723

Field Detail

y
```
protected DoubleMatrix y
```

lambda
```
protected kotlin.Array[] lambda
```

eta2
```
protected kotlin.Array[] eta2
```

c
```
protected double c
```

N
```
protected int N
```

q
```
protected int q
```

Nran
```
protected int Nran
```

dfMarker
```
protected int dfMarker
```

Xoriginal
```
protected DoubleMatrix Xoriginal
```

X
```
protected DoubleMatrix X
```

Zoriginal
```
protected DoubleMatrix Zoriginal
```

Z
```
protected DoubleMatrix Z
```

K
```
protected DoubleMatrix K
```

eig
```
protected EigenvalueDecomposition eig
```

eigA
```
protected EigenvalueDecomposition eigA
```

U
```
protected DoubleMatrix U
```

invH
```
protected DoubleMatrix invH
```

invXHX
```
protected DoubleMatrix invXHX
```

beta
```
protected DoubleMatrix beta
```

Xbeta
```
protected DoubleMatrix Xbeta
```

ssModel
```
protected double ssModel
```

ssError
```
protected double ssError
```

SST
```
protected double SST
```

Rsq
```
protected double Rsq
```

dfModel
```
protected int dfModel
```

dfError
```
protected int dfError
```

delta
```
protected double delta
```

varResidual
```
protected double varResidual
```

varRandomEffect
```
protected double varRandomEffect
```

blup
```
protected DoubleMatrix blup
```

pred
```
protected DoubleMatrix pred
```

res
```
protected DoubleMatrix res
```

pev
```
protected DoubleMatrix pev
```

lnLikelihood
```
protected double lnLikelihood
```

findDelta
```
protected boolean findDelta
```

calculatePEV
```
protected boolean calculatePEV
```

lowerlimit
```
protected double lowerlimit
```

upperlimit
```
protected double upperlimit
```

nregions
```
protected int nregions
```

convergence
```
protected double convergence
```

maxiter
```
protected int maxiter
```

subintervalCount
```
protected int subintervalCount
```

Constructor Detail

EMMAforDoubleMatrix

public EMMAforDoubleMatrix(DoubleMatrix y,
                           DoubleMatrix fixed,
                           DoubleMatrix kin,
                           int nAlleles)

EMMAforDoubleMatrix
```
public EMMAforDoubleMatrix(DoubleMatrix data,
                           DoubleMatrix fixed,
                           DoubleMatrix kin,
                           int nAlleles,
                           double delta)
```
This constructor assumes that Z is the identity matrix for calculating blups, predicted values and residuals. If that is not true use the contstructor that explicity takes Z. This constructor treats A as ZKZ' so it can be used if blups and residuals are not needed.

Parameters:

data -

fixed -

kin -

nAlleles -

delta -

EMMAforDoubleMatrix

public EMMAforDoubleMatrix(DoubleMatrix data,
                           DoubleMatrix fixed,
                           DoubleMatrix kin,
                           DoubleMatrix inZ,
                           int nAlleles,
                           double delta)

This constructor should be used when Z is not the identity matrix. Z is needed to calculate blups and residuals.

Parameters:: data -; fixed -; kin -; inZ -; nAlleles -; delta -

EMMAforDoubleMatrix
```
public EMMAforDoubleMatrix(DoubleMatrix data,
                           DoubleMatrix fixed,
                           DoubleMatrix kin)
```
This constructor is designed for G-BLUP. Uses kinship matrix with phenotypes in same order. Phenotypes = NaN for prediction set. Assumes that Z is the identity matrix for calculating blups, predicted values and residuals. If that is not true use the constructor that explicity takes Z. This constructor treats A as ZKZ' so it can be used if blups and residuals are not needed.

Parameters:

data - A column double matrix of phenotypic values (number of individuals x 1 taxa ID column + number of traits)

fixed -

kin -

Method Detail

init
```
protected void init()
```

solve
```
public void solve()
```

calculateBlupsPredictedResiduals

public void calculateBlupsPredictedResiduals()

calculateBlupsPredicted
```
public void calculateBlupsPredicted()
```

calculateBLUP
```
public void calculateBLUP()
```

getDfMarker
```
public int getDfMarker()
```

getBeta
```
public DoubleMatrix getBeta()
```

getDfModel
```
public int getDfModel()
```

getDfError
```
public int getDfError()
```

getDelta
```
public double getDelta()
```

getInvH
```
public DoubleMatrix getInvH()
```

getVarRes
```
public double getVarRes()
```

getVarRan
```
public double getVarRan()
```

getBlup
```
public DoubleMatrix getBlup()
```

getPev
```
public DoubleMatrix getPev()
```

getPred
```
public DoubleMatrix getPred()
```

getRes
```
public DoubleMatrix getRes()
```

getLnLikelihood
```
public double getLnLikelihood()
```

getMarkerFp
```
public kotlin.Array[] getMarkerFp()
```
Returns:

org.jetbrains.dokka.NodeRenderContent@6b943294: For markers with 2 df, F.fullModel, p.fullModel, additive effect, Fadd, padd, dominance effect, Fdom, pdom For markers with other than 2 df, F and p for the full model only

solveWithNewData

public void solveWithNewData(DoubleMatrix y)

setCalculatePEV

public void setCalculatePEV(boolean calculatePEV)

Class EMMAforDoubleMatrix

Field Detail

y

lambda

eta2

c

N

q

Nran

dfMarker

Xoriginal

X

Zoriginal

Z

K

eig

eigA

U

invH

invXHX

beta

Xbeta

ssModel

ssError

SST

Rsq

dfModel

dfError

delta

varResidual

varRandomEffect

blup

pred

res

pev

lnLikelihood

findDelta

calculatePEV

lowerlimit

upperlimit

nregions

convergence

maxiter

subintervalCount

Constructor Detail

EMMAforDoubleMatrix

EMMAforDoubleMatrix

EMMAforDoubleMatrix

EMMAforDoubleMatrix

Method Detail

init

solve

calculateBlupsPredictedResiduals

calculateBlupsPredicted

calculateBLUP

getDfMarker

getBeta

getDfModel

getDfError

getDelta

getInvH

getVarRes

getVarRan

getBlup

getPev

getPred

getRes

getLnLikelihood

getMarkerFp

solveWithNewData

setCalculatePEV