no.uib.cipr.matrix.io.MatrixVectorReader Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of mt-java Show documentation
Show all versions of mt-java Show documentation
Matrix data structures, linear solvers, least squares methods, eigenvalue,
and singular value decompositions. For larger random dense matrices (above ~ 350 x 350)
matrix-matrix multiplication C = A.B is about 50% faster than MTJ.
/*
* Copyright (C) 2003-2006 Bjørn-Ove Heimsund
*
* This file is part of MTJ.
*
* This library is free software; you can redistribute it and/or modify it
* under the terms of the GNU Lesser General Public License as published by the
* Free Software Foundation; either version 2.1 of the License, or (at your
* option) any later version.
*
* This library is distributed in the hope that it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License
* for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this library; if not, write to the Free Software Foundation,
* Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
package no.uib.cipr.matrix.io;
import java.io.BufferedReader;
import java.io.EOFException;
import java.io.IOException;
import java.io.Reader;
import java.io.StreamTokenizer;
import java.util.LinkedList;
import java.util.List;
/**
* Reads matrices and vectors
*/
public class MatrixVectorReader extends BufferedReader {
/**
* Reads the entries of the matrix or vector
*/
private StreamTokenizer st;
/**
* Constructor for MatrixVectorReader
*
* @param in
* A Reader
*/
public MatrixVectorReader(Reader in) {
super(in);
setup();
}
/**
* Constructor for MatrixVectorReader
*
* @param in
* A Reader
* @param sz
* Input buffer size
*/
public MatrixVectorReader(Reader in, int sz) {
super(in, sz);
setup();
}
/**
* Sets up the stream tokenizer
*/
private void setup() {
st = new StreamTokenizer(this);
st.resetSyntax();
st.eolIsSignificant(false);
st.lowerCaseMode(true);
// Parse numbers as words
st.wordChars('0', '9');
st.wordChars('-', '.');
// Characters as words
st.wordChars('\u0000', '\u00FF');
// Skip comments
st.commentChar('%');
// Skip whitespace and newlines
st.whitespaceChars(' ', ' ');
st.whitespaceChars('\u0009', '\u000e');
}
/**
* Shifts the indices. Useful for converting between 0- and 1-based
* indicing.
*
* @param num
* Added to every index
* @param indices
* Indices to shift
*/
public void add(int num, int[] indices) {
for (int i = 0; i < indices.length; ++i)
indices[i] += num;
}
/**
* Reads a line, and trims it of surrounding whitespace
*
* @throws IOException
* If either I/O errors occur, or there was nothing to read
*/
private String readTrimmedLine() throws IOException {
String line = readLine();
if (line != null)
return line.trim();
else
throw new EOFException();
}
/**
* Reads the matrix info for the Matrix Market exchange format. The line
* must consist of exactly 5 space-separated entries, the first being
* "%%MatrixMarket"
*/
public MatrixInfo readMatrixInfo() throws IOException {
String[] component = readTrimmedLine().split(" +");
if (component.length != 5)
throw new IOException(
"Current line unparsable. It must consist of 5 tokens");
// Read header
if (!component[0].equalsIgnoreCase("%%MatrixMarket"))
throw new IOException("Not in Matrix Market exchange format");
// This will always be "matrix"
if (!component[1].equalsIgnoreCase("matrix"))
throw new IOException("Expected \"matrix\", got " + component[1]);
// Sparse or dense?
boolean sparse = false;
if (component[2].equalsIgnoreCase("coordinate"))
sparse = true;
else if (component[2].equalsIgnoreCase("array"))
sparse = false;
else
throw new IOException("Unknown layout " + component[2]);
// Dataformat
MatrixInfo.MatrixField field = null;
if (component[3].equalsIgnoreCase("real"))
field = MatrixInfo.MatrixField.Real;
else if (component[3].equalsIgnoreCase("integer"))
field = MatrixInfo.MatrixField.Integer;
else if (component[3].equalsIgnoreCase("complex"))
field = MatrixInfo.MatrixField.Complex;
else if (component[3].equalsIgnoreCase("pattern"))
field = MatrixInfo.MatrixField.Pattern;
else
throw new IOException("Unknown field specification " + component[3]);
// Matrix pattern
MatrixInfo.MatrixSymmetry symmetry = null;
if (component[4].equalsIgnoreCase("general"))
symmetry = MatrixInfo.MatrixSymmetry.General;
else if (component[4].equalsIgnoreCase("symmetric"))
symmetry = MatrixInfo.MatrixSymmetry.Symmetric;
else if (component[4].equalsIgnoreCase("skew-symmetric"))
symmetry = MatrixInfo.MatrixSymmetry.SkewSymmetric;
else if (component[4].equalsIgnoreCase("Hermitian"))
symmetry = MatrixInfo.MatrixSymmetry.Hermitian;
else
throw new IOException("Unknown symmetry specification "
+ component[4]);
// Pack together. This also verifies the format
return new MatrixInfo(sparse, field, symmetry);
}
/**
* Reads the vector info for the Matrix Market exchange format. The line
* must consist of exactly 4 space-separated entries, the first being
* "%%MatrixMarket"
*/
public VectorInfo readVectorInfo() throws IOException {
String[] component = readTrimmedLine().split(" +");
if (component.length != 4)
throw new IOException(
"Current line unparsable. It must consist of 4 tokens");
// Read header
if (!component[0].equalsIgnoreCase("%%MatrixMarket"))
throw new IOException("Not in Matrix Market exchange format");
// This will always be "vector"
if (!component[1].equalsIgnoreCase("vector"))
throw new IOException("Expected \"vector\", got " + component[1]);
// Sparse or dense?
boolean sparse = false;
if (component[2].equalsIgnoreCase("coordinate"))
sparse = true;
else if (component[2].equalsIgnoreCase("array"))
sparse = false;
else
throw new IOException("Unknown layout " + component[2]);
// Dataformat
VectorInfo.VectorField field = null;
if (component[3].equalsIgnoreCase("real"))
field = VectorInfo.VectorField.Real;
else if (component[3].equalsIgnoreCase("integer"))
field = VectorInfo.VectorField.Integer;
else if (component[3].equalsIgnoreCase("complex"))
field = VectorInfo.VectorField.Complex;
else if (component[3].equalsIgnoreCase("pattern"))
field = VectorInfo.VectorField.Pattern;
else
throw new IOException("Unknown field specification " + component[3]);
// Pack together. This also verifies the format
return new VectorInfo(sparse, field);
}
/**
* Checks if a Matrix Market header is present ("%%MatrixMarket")
*
* @return True if a header was found, else false
* @throws IOException
*/
public boolean hasInfo() throws IOException {
// Read a line, then skip back
mark(1024);
String[] component = readTrimmedLine().split(" +");
reset();
return component[0].equalsIgnoreCase("%%MatrixMarket");
}
/**
* Reads all the comments (lines starting with '%'). Positions the reader at
* the first non-comment line. Can only be called after reading the matrix
* or vector info. The comments read does not include '%' or the newline
*/
public String[] readComments() throws IOException {
List list = new LinkedList();
while (true) {
mark(1024); // Line length equal 1024 at most
String line = readTrimmedLine();
if (line.length() > 0)
if (line.charAt(0) != '%') {
reset();
break;
} else
list.add(line.substring(1));
}
return list.toArray(new String[list.size()]);
}
/**
* Reads in the size of a matrix. Skips initial comments
*/
public MatrixSize readMatrixSize(MatrixInfo info) throws IOException {
// Always read the matrix size
int numRows = getInt(), numColumns = getInt();
// For coordinate matrices we also read the number of entries
if (info.isDense())
return new MatrixSize(numRows, numColumns, info);
else {
int numEntries = getInt();
return new MatrixSize(numRows, numColumns, numEntries);
}
}
/**
* Reads in the size of an array matrix. Skips initial comments
*/
public MatrixSize readArraySize() throws IOException {
int numRows = getInt(), numColumns = getInt();
return new MatrixSize(numRows, numColumns, numRows * numColumns);
}
/**
* Reads in the size of a coordinate matrix. Skips initial comments
*/
public MatrixSize readCoordinateSize() throws IOException {
int numRows = getInt(), numColumns = getInt(), numEntries = getInt();
return new MatrixSize(numRows, numColumns, numEntries);
}
/**
* Reads in the size of a vector. Skips initial comments
*/
public VectorSize readVectorSize(VectorInfo info) throws IOException {
// Always read the vector size
int size = getInt();
// For coordinate vectors we also read the number of entries
if (info.isDense())
return new VectorSize(size);
else {
int numEntries = getInt();
return new VectorSize(size, numEntries);
}
}
/**
* Reads in the size of a dense vector. Skips initial comments
*/
public VectorSize readVectorArraySize() throws IOException {
int size = getInt();
return new VectorSize(size);
}
/**
* Reads in the size of a coordinate vector. Skips initial comments
*/
public VectorSize readVectorCoordinateSize() throws IOException {
int size = getInt(), numEntries = getInt();
return new VectorSize(size, numEntries);
}
/**
* Reads the array data
*/
public void readArray(double[] data) throws IOException {
int size = data.length;
for (int i = 0; i < size; ++i)
data[i] = getDouble();
}
/**
* Reads the array data
*/
public void readArray(float[] data) throws IOException {
int size = data.length;
for (int i = 0; i < size; ++i)
data[i] = getFloat();
}
/**
* Reads the array data
*/
public void readArray(int[] data) throws IOException {
int size = data.length;
for (int i = 0; i < size; ++i)
data[i] = getInt();
}
/**
* Reads the array data
*/
public void readArray(long[] data) throws IOException {
int size = data.length;
for (int i = 0; i < size; ++i)
data[i] = getLong();
}
/**
* Reads the array data. The first array will contain real entries, while
* the second contain imaginary entries
*/
public void readArray(double[] dataR, double[] dataI) throws IOException {
int size = dataR.length;
if (size != dataI.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
dataR[i] = getDouble();
dataI[i] = getDouble();
}
}
/**
* Reads the array data. The first array will contain real entries, while
* the second contain imaginary entries
*/
public void readArray(float[] dataR, float[] dataI) throws IOException {
int size = dataR.length;
if (size != dataI.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
dataR[i] = getFloat();
dataI[i] = getFloat();
}
}
/**
* Reads a coordinate vector
*/
public void readCoordinate(int[] index, double[] data) throws IOException {
int size = index.length;
if (size != data.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
index[i] = getInt();
data[i] = getDouble();
}
}
/**
* Reads a coordinate vector
*/
public void readCoordinate(int[] index, float[] data) throws IOException {
int size = index.length;
if (size != data.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
index[i] = getInt();
data[i] = getFloat();
}
}
/**
* Reads a coordinate vector
*/
public void readCoordinate(int[] index, int[] data) throws IOException {
int size = index.length;
if (size != data.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
index[i] = getInt();
data[i] = getInt();
}
}
/**
* Reads a coordinate vector
*/
public void readCoordinate(int[] index, long[] data) throws IOException {
int size = index.length;
if (size != data.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
index[i] = getInt();
data[i] = getLong();
}
}
/**
* Reads a coordinate vector. First data array contains real entries, and
* the second contains imaginary entries
*/
public void readCoordinate(int[] index, float[] dataR, float[] dataI)
throws IOException {
int size = index.length;
if (size != dataR.length || size != dataI.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
index[i] = getInt();
dataR[i] = getFloat();
dataI[i] = getFloat();
}
}
/**
* Reads a coordinate vector. First data array contains real entries, and
* the second contains imaginary entries
*/
public void readCoordinate(int[] index, double[] dataR, double[] dataI)
throws IOException {
int size = index.length;
if (size != dataR.length || size != dataI.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
index[i] = getInt();
dataR[i] = getDouble();
dataI[i] = getDouble();
}
}
/**
* Reads a pattern vector
*/
public void readPattern(int[] index) throws IOException {
int size = index.length;
for (int i = 0; i < size; ++i)
index[i] = getInt();
}
/**
* Reads a coordinate matrix
*/
public void readCoordinate(int[] row, int[] column, double[] data)
throws IOException {
int size = row.length;
if (size != column.length || size != data.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
row[i] = getInt();
column[i] = getInt();
data[i] = getDouble();
}
}
/**
* Reads a coordinate matrix
*/
public void readCoordinate(int[] row, int[] column, float[] data)
throws IOException {
int size = row.length;
if (size != column.length || size != data.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
row[i] = getInt();
column[i] = getInt();
data[i] = getFloat();
}
}
/**
* Reads a coordinate matrix
*/
public void readCoordinate(int[] row, int[] column, int[] data)
throws IOException {
int size = row.length;
if (size != column.length || size != data.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
row[i] = getInt();
column[i] = getInt();
data[i] = getInt();
}
}
/**
* Reads a coordinate matrix
*/
public void readCoordinate(int[] row, int[] column, long[] data)
throws IOException {
int size = row.length;
if (size != column.length || size != data.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
row[i] = getInt();
column[i] = getInt();
data[i] = getLong();
}
}
/**
* Reads a pattern matrix
*/
public void readPattern(int[] row, int[] column) throws IOException {
int size = row.length;
if (size != column.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
row[i] = getInt();
column[i] = getInt();
}
}
/**
* Reads a coordinate matrix. First data array contains real entries, and
* the second contains imaginary entries
*/
public void readCoordinate(int[] row, int[] column, double[] dataR,
double[] dataI) throws IOException {
int size = row.length;
if (size != column.length || size != dataR.length
|| size != dataI.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
row[i] = getInt();
column[i] = getInt();
dataR[i] = getDouble();
dataI[i] = getDouble();
}
}
/**
* Reads a coordinate matrix. First data array contains real entries, and
* the second contains imaginary entries
*/
public void readCoordinate(int[] row, int[] column, float[] dataR,
float[] dataI) throws IOException {
int size = row.length;
if (size != column.length || size != dataR.length
|| size != dataI.length)
throw new IllegalArgumentException(
"All arrays must be of the same size");
for (int i = 0; i < size; ++i) {
row[i] = getInt();
column[i] = getInt();
dataR[i] = getFloat();
dataI[i] = getFloat();
}
}
/**
* Reads an integer
*/
private int getInt() throws IOException {
st.nextToken();
if (st.ttype == StreamTokenizer.TT_WORD)
return Double.valueOf(st.sval).intValue();
else if (st.ttype == StreamTokenizer.TT_EOF)
throw new EOFException("End-of-File encountered during parsing");
else
throw new IOException("Unknown token found during parsing");
}
/**
* Reads a long
*/
private long getLong() throws IOException {
st.nextToken();
if (st.ttype == StreamTokenizer.TT_WORD)
return Long.parseLong(st.sval);
else if (st.ttype == StreamTokenizer.TT_EOF)
throw new EOFException("End-of-File encountered during parsing");
else
throw new IOException("Unknown token found during parsing");
}
/**
* Reads a double
*/
private double getDouble() throws IOException {
st.nextToken();
if (st.ttype == StreamTokenizer.TT_WORD)
return Double.parseDouble(st.sval);
else if (st.ttype == StreamTokenizer.TT_EOF)
throw new EOFException("End-of-File encountered during parsing");
else
throw new IOException("Unknown token found during parsing");
}
/**
* Reads a float
*/
private float getFloat() throws IOException {
st.nextToken();
if (st.ttype == StreamTokenizer.TT_WORD)
return Float.parseFloat(st.sval);
else if (st.ttype == StreamTokenizer.TT_EOF)
throw new EOFException("End-of-File encountered during parsing");
else
throw new IOException("Unknown token found during parsing");
}
}
© 2015 - 2024 Weber Informatics LLC | Privacy Policy