JMathTools is a collection of independent packages designed to fit common engineering/scientific computing needs. For now, JMathTools contains:

• JMathArray: simple linear algebra for double[][] arrays using static methods
• JMathPlot: interactive plot in 2D / 3D
• JMathIO: binary and ASCII input/output for double arrays

JMathTools is designed to be:

• easy to integrate in your java applications
• easy to extend and modify

It was initially developed to provide easy to use java API in Matlab to Java porting task.

JMathTools was NOT designed to be:

• a high performance computing engine
• a high performance 3D visualization tool

For such things, many free Java packages are available (see Links).

Provides a list of static methods applicable on int and double arrays:

• merge, insert, delete rows and columns : org.math.array.DoubleArray,org.math.array.IntegerArray
• simple linear algebra: LU, QR, EigenValues, Cholesky, SingularValues (based on JAMA package): org.math.array.LinearAlgebra
• simple statistics: mean, variance, covariance, histograms : org.math.array.StatisticSample
• some random generators (based on RnGPack): normal, lognormal, triangular, chi square, weibull, ... : org.math.array.StatisticSample

import static org.math.array.LinearAlgebra.*;
...
  // random 4 x 4 matrix + Id
  double[][] A = plus(random(4, 4), identity(4));
		
  // Eigen values Decomposition : A = V * D * V^-1
  EigenvalueDecomposition eig = eigen(A);
  double[][] V = eig.getV();
  double[][] D = eig.getD();

JMathArray syntax is stable. Only new features may be added, and current syntax will stay available.

→ Tutorial

Provides :

• 2D/3D scatter plot
• 2D/3D line plot
• 2D staircase plot
• 2D/3D histogram plot
• 2D/3D boxplot
• 3D grid plot
• 2D/3D quantiles on plots

import org.math.plot.*;
...
  
  double[] x = ...
  double[] y = ...
 
  // create your PlotPanel (you can use it as a JPanel)
  Plot2DPanel plot = new Plot2DPanel();
 
  // add a line plot to the PlotPanel
  plot.addLinePlot("my plot", x, y);
 
  // put the PlotPanel in a JFrame, as a JPanel
  JFrame frame = new JFrame("a plot panel");
  frame.setContentPane(plot);
  frame.setVisible(true);

JMathPlot syntax is NOT stable. As far as possible current syntax will stay available.

→ Tutorial

JMathIO allows to create, read, write, append your datas in:

• ASCII files : org.math.io.files.ASCIIFile
• Binary files (big or little endian) : org.math.io.files.BinaryFile

import static org.math.io.files.ASCIIFile.*;
import static org.math.io.parser.ArrayString.*;
 
...
 
  double[][] b = { { 1.0, 2.0, 3.0, 4.0 }, { 11.0, 12.0, 13.0, 14.0 } };
  // write file in ASCII
  writeDoubleArray(new File("b.txt"), b);
  // read this file
  double[][] b_read = readDoubleArray(new File("b.txt"));

JMathIO syntax is considered stable. Only new features may be added, and current syntax will stay available.

→ Tutorial

GroovyLab is a full jmathtools wrapper for Groovy agile dynamic language (close to Ruby, Python and Java). It is a collection of Groovy classes to provide common math features like math enginneering languages (matlab, scilab, ...):

• Matrix class
• Plot class

Basically, it allows to create scripts with matlab-like syntax for Matrices and plots:

import static org.math.array.Matrix.*
import static org.math.plot.Plot.*
 
def A = rand(10,3) 		// random Matrix of 10 rows and 3 columns
def B = fill(10,3,1.0) 	// one Matrix of 10 rows and 3 columns
def C = A + B 			// support for matrix addition with "+" or "-"
def D = A - 2.0			// support for number addition with "+" or "-"
def E = A * B			// support for matrix multiplication or division
def F = rand(3,3)	
def G = F**(-1)			// support for matrix power (with integers only)
 
println A				// display Matrix content
 
plot("A",A,"SCATTER")	// plot Matrix values as ScatterPlot
 
def M = rand(5,5) + id(5) //Eigenvalues decomposition
println "M=\n" + M
println "V=\n" + V(M)
println "D=\n" + D(M)
println "M~\n" + (V(M) * D(M) * V(M)**(-1))
 
...
def A = rand(10,3)
println A
plot("A",A,"SCATTER")
 
def B = rand(10,3)
println B
plot("B",B,"LINE")
...

Rsession is an easy to use java class providing access to remote or local R session. The back-end engine is Rserve 0.6, locally spawned automatically if necessary.

import static org.math.R.Rsession.*;
...
 
    public static void main(String args[]) {
        Rsession s = Rsession.newLocalInstance(System.out);
 
        double[] rand=((double[]) cast(s.eval("rnorm(10)"))); //create java variable from R command
 
        ...
 
        s.set("c", Math.random()); //create R variable from java one
 
        s.save(new File("save.Rdata"), "c"); //save variables in .Rdata
        s.rm("c"); //delete variable in R environment
        s.load(new File("save.Rdata")); //load R variable from .Rdata
 
        ...
 
        s.set("df", new double[][]{{1, 2, 3}, {4, 5, 6}, {7, 8, 9}, {10, 11, 12}}, "x1", "x2", "x3"); //create data frame from given vectors
        double value= (Double) (cast(s.eval("df$x1[3]"))); //access one value in data frame
 
        ...
 
        s.toJPEG(new File("plot.jpg"), 400, 400, "plot(rnorm(10))"); //create jpeg file from R graphical command (like plot)
 
        String html = s.asHTML("summary(rnorm(100))"); //format in html using R2HTML
        System.out.println(html);
 
        String txt = s.asString("summary(rnorm(100))"); //format in text
        System.out.println(txt);
 
        ...
 
        System.out.println(s.installPackage("sensitivity", true)); //install and load R package
        System.out.println(s.installPackage("wavelets", true));
 
        s.end();
    }