X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=linearalgebra.h;h=7453f4ebd8b86ed20f530c2e1ba7989808361fb2;hb=3e8b80da722e11c72bce957e2f42a6e884dd02b6;hp=58a60dae8f04d9eae12fff9f8d729cc2119d4c92;hpb=f0a594f6676ef5a52d1f122b6de70de2fda08c81;p=mothur.git

diff --git a/linearalgebra.h b/linearalgebra.h
index 58a60da..7453f4e 100644
--- a/linearalgebra.h
+++ b/linearalgebra.h
@@ -20,20 +20,63 @@ public:
 	~LinearAlgebra() {}
 	
 	vector<vector<double> > matrix_mult(vector<vector<double> >, vector<vector<double> >);
+    vector<vector<double> >transpose(vector<vector<double> >);
 	void recenter(double, vector<vector<double> >, vector<vector<double> >&);
-	int tred2(vector<vector<double> >&, vector<double>&, vector<double>&);
+	//eigenvectors
+    int tred2(vector<vector<double> >&, vector<double>&, vector<double>&);
 	int qtli(vector<double>&, vector<double>&, vector<vector<double> >&);
+    
 	vector< vector<double> > calculateEuclidianDistance(vector<vector<double> >&, int); //pass in axes and number of dimensions
 	vector< vector<double> > calculateEuclidianDistance(vector<vector<double> >&); //pass in axes
 	vector<vector<double> > getObservedEuclideanDistance(vector<vector<double> >&);
 	double calcPearson(vector<vector<double> >&, vector<vector<double> >&);
 	double calcSpearman(vector<vector<double> >&, vector<vector<double> >&);
 	double calcKendall(vector<vector<double> >&, vector<vector<double> >&);
-
+    double calcKruskalWallis(vector<spearmanRank>&, double&);
+    double calcWilcoxon(vector<double>&, vector<double>&, double&);
+	
+	double calcPearson(vector<double>&, vector<double>&, double&);
+	double calcSpearman(vector<double>&, vector<double>&, double&);
+	double calcKendall(vector<double>&, vector<double>&, double&);
+    
+	double calcSpearmanSig(double, double, double, double); //length, f^3 - f where f is the number of ties in x, f^3 - f where f is the number of ties in y, sum of squared diffs in ranks. - designed to find the sif of one score.
+    double calcPearsonSig(double, double); //length, coeff.
+    double calcKendallSig(double, double); //length, coeff.
+    
+    vector<double> solveEquations(vector<vector<double> >, vector<double>);
+    vector<float> solveEquations(vector<vector<float> >, vector<float>);
+    vector<vector<double> > getInverse(vector<vector<double> >);
+    double choose(double, double);
+    double normalvariate(double mu, double sigma);
+    vector< vector<double> > lda(vector< vector<double> >& a, vector<string> groups, vector< vector<double> >& means, bool&); //Linear discriminant analysis - a is [features][valuesFromGroups] groups indicates which group each sampling comes from. For example if groups = early, late, mid, early, early. a[0][0] = value for feature0 from groupEarly.
+    int svd(vector< vector<double> >& a, vector<double>& w, vector< vector<double> >& v); //Singular value decomposition
 private:
 	MothurOut* m;
 	
 	double pythag(double, double);
+    double betacf(const double, const double, const double);
+    double betai(const double, const double, const double);
+    double gammln(const double);
+    double gammq(const double, const double);
+    double gser(double&, const double, const double, double&);
+    double gcf(double&, const double, const double, double&);
+    double erfcc(double);
+    double gammp(const double, const double);
+    double pnorm(double x);
+    
+    double ran0(int&); //for testing 
+    double ran1(int&); //for testing
+    double ran2(int&); //for testing
+    double ran3(int&); //for testing
+    double ran4(int&); //for testing
+    void psdes(unsigned long &, unsigned long &); //for testing
+    
+    void ludcmp(vector<vector<double> >&, vector<int>&, double&);
+    void lubksb(vector<vector<double> >&, vector<int>&, vector<double>&);
+    
+    void ludcmp(vector<vector<float> >&, vector<int>&, float&);
+    void lubksb(vector<vector<float> >&, vector<int>&, vector<float>&);
+    
 };
 
 #endif