NMath User's Guide

TOC | Previous | Next | Index

44.2 One-Way Repeated Measures ANOVA (.NET, C#, CSharp, VB, Visual Basic, F#)

Class OneWayRanova calculates and summarizes the information of a one-way repeated measures analysis of variance (RANOVA).

Creating One-Way RANOVA Objects

A OneWayRanova instance is constructed from numeric data for multiple treatments applied to each experimental subject. For example, this code constructs a OneWayRanova from a DoubleMatrix:

Code Example – C# RANOVA

var data = new DoubleMatrix( "8x4 [ 180 200 160 200

                                    230 250 200 220

                                    280 310 260 270

                                    180 200 160 200

                                    190 210 170 210

                                    140 160 120 110

                                    270 300 250 260 

                                    110 130 100 100 ]" );

var ranova = new OneWayRanova( data );

Code Example – VB RANOVA

Dim Data As New DoubleMatrix("8x4 [ 180 200 160 200

                                    230 250 200 220

                                    280 310 260 270

                                    180 200 160 200

                                    190 210 170 210

                                    140 160 120 110

                                    270 300 250 260 

                                    110 130 100 100 ]")

Dim Ranova As New OneWayRanova(Data)

Each row of the matrix contains the data for an individual subject. There should be one column for each treatment. The example above shows 4 different measurements for each of 8 subjects.

NOTE—Data rows containing missing values (NaNs) are ignored by class OneWayRanova.

Similarly, you can also construct a OneWayRanova from a DataFrame:

Code Example – C# RANOVA

var ranova = new OneWayRanova( df );

Code Example – VB RANOVA

Dim Ranova As New OneWayRanova(DF)

Each row in the DataFrame contains the data for an individual subject. There should be one column for each treatment.

Note that all numeric columns in the given DataFrame are interpreted as treatments; only non-numeric columns are ignored. If you have numeric columns in the data frame that you also wish to ignore, apply the appropriate Subset first. For instance:

Code Example – C# RANOVA

var colIndices = new Subset( new int[] { 3, 14, 5, 8, 4 } );

var ranova = new OneWayRanova( df.GetColumns( colIndices ) );

Code Example – VB RANOVA

Dim ColIndices As New Subset(New Integer() {3, 14, 5, 8, 4})

Dim Ranova As New OneWayRanova(DF.GetColumns(ColIndices))

The One-Way RANOVA Table

Once you've constructed a OneWayRanova, you can display the complete RANOVA table:

Code Example – C# RANOVA

Console.WriteLine( ranova );

Code Example – VB RANOVA

Console.WriteLine(Ranova)

For example:




Source   Deg of Freedom  Sum Of Sq   Mean Square     F       P




Subjects      9         102822.5000  11424.7222    .        .




Treatment     3         9247.5000    3082.5000     31.6755  0.0000




Error        27         2627.5000    97.3148       .        .




Total        39         114697.5000  2940.9615     .        .




Class OneWayRanovaTable is provided
for summarizing the information in a traditional one-way RANOVA table.
Class OneWayRanovaTable derives
from DataFrame. An instance of
OneWayRanovaTable can be obtained
from a OneWayRanova object using
the RanovaTable property. For example:


Code Example – C# RANOVA


OneWayRanovaTable myTable = ranova.RanovaTable;




Code Example – VB RANOVA


Dim MyTable As OneWayRanovaTable = Ranova.RanovaTable




Class OneWayRanovaTable
provides the following read-only properties for accessing individual
elements in the RANOVA table:


● DegreesOfFreedomTreatment gets the treatment degrees
of freedom.


● DegreesOfFreedomWithinSubject gets the within-subject
degrees of freedom.


● DegreesOfFreedomError gets the error degrees of
freedom.


● DegreesOfFreedomTotal gets the total degrees of
freedom.


● SumOfSquaresTreatment gets the treatment sum of
squares.


● SumOfSquaresWithinSubject gets the within-subject
sum of squares.


● SumOfSquaresTotal gets the total sum of squares.


● SumOfSquaresError gets the error sum of squares.


● MeanSquareTreatment gets the treatment mean square.


● MeanSquareWithinSubject gets the within-subject
mean square.


● MeanSquareError gets the error mean square.


● MeanSquareTotal gets the total mean square.


● FStatistic gets the F statistic for the RANOVA.


● FStatisticPValue gets the p-value for the F statistic.


Grand Mean, Subject Means, and Treatment Means


Class OneWayRanova
provides properties for retrieving the grand mean, subject means, and treatment means:


● GrandMean gets the grand mean of the data. The
grand mean is the mean of all of the data.


● SubjectMeans gets a vector of means for each subject.


● TreatmentMeans gets a vector of means for each
treatment.


Critical Value of the F Statistic


Class OneWayRanova
provides the convenience function FStatisticCriticalValue()
which computes the critical value for the RANOVA F statistic at a
given significance level. Thus:


Code Example – C# RANOVA


double alpha = 0.01;




double critVal = ranova.FStatisticCriticalValue( alpha );




Code Example – VB RANOVA


Dim Alpha As Double = 0.01




Dim CritVal As Double = Ranova.FStatisticCriticalValue(Alpha)




Updating One-Way RANOVA Objects


Method SetData() updates
an entire repeated measures analysis of variance object with new data.
As with the class constructors (see above), you can supply data as a
matrix or as a data frame. For instance, this code updates a RANOVA with
data from matrix A:


Code Example – C# RANOVA


ranova.SetData( A );




Code Example – VB RANOVA


Ranova.SetData(A)







Top

		


		
	

		

	


	



Top