code
stringlengths
73
34.1k
label
stringclasses
1 value
public void growMaxLength( int arrayLength , boolean preserveValue ) { if( arrayLength < 0 ) throw new IllegalArgumentException("Negative array length. Overflow?"); // see if multiplying numRows*numCols will cause an overflow. If it won't then pick the smaller of the two if( numRows != 0 && numCols <= Integer.MAX_VALUE / numRows ) { // save the user from themselves arrayLength = Math.min(numRows*numCols, arrayLength); } if( nz_values == null || arrayLength > this.nz_values.length ) { double[] data = new double[ arrayLength ]; int[] row_idx = new int[ arrayLength ]; if( preserveValue ) { if( nz_values == null ) throw new IllegalArgumentException("Can't preserve values when uninitialized"); System.arraycopy(this.nz_values, 0, data, 0, this.nz_length); System.arraycopy(this.nz_rows, 0, row_idx, 0, this.nz_length); } this.nz_values = data; this.nz_rows = row_idx; } }
java
public void growMaxColumns( int desiredColumns , boolean preserveValue ) { if( col_idx.length < desiredColumns+1 ) { int[] c = new int[ desiredColumns+1 ]; if( preserveValue ) System.arraycopy(col_idx,0,c,0,col_idx.length); col_idx = c; } }
java
public void histogramToStructure(int histogram[] ) { col_idx[0] = 0; int index = 0; for (int i = 1; i <= numCols; i++) { col_idx[i] = index += histogram[i-1]; } nz_length = index; growMaxLength( nz_length , false); if( col_idx[numCols] != nz_length ) throw new RuntimeException("Egads"); }
java
public void sortIndices(SortCoupledArray_F64 sorter ) { if( sorter == null ) sorter = new SortCoupledArray_F64(); sorter.quick(col_idx,numCols+1,nz_rows,nz_values); indicesSorted = true; }
java
public void copyStructure( DMatrixSparseCSC orig ) { reshape(orig.numRows, orig.numCols, orig.nz_length); this.nz_length = orig.nz_length; System.arraycopy(orig.col_idx,0,col_idx,0,orig.numCols+1); System.arraycopy(orig.nz_rows,0,nz_rows,0,orig.nz_length); }
java
public static boolean bidiagOuterBlocks( final int blockLength , final DSubmatrixD1 A , final double gammasU[], final double gammasV[]) { // System.out.println("---------- Orig"); // A.original.print(); int width = Math.min(blockLength,A.col1-A.col0); int height = Math.min(blockLength,A.row1-A.row0); int min = Math.min(width,height); for( int i = 0; i < min; i++ ) { //--- Apply reflector to the column // compute the householder vector if (!computeHouseHolderCol(blockLength, A, gammasU, i)) return false; // apply to rest of the columns in the column block rank1UpdateMultR_Col(blockLength,A,i,gammasU[A.col0+i]); // apply to the top row block rank1UpdateMultR_TopRow(blockLength,A,i,gammasU[A.col0+i]); System.out.println("After column stuff"); A.original.print(); //-- Apply reflector to the row if(!computeHouseHolderRow(blockLength,A,gammasV,i)) return false; // apply to rest of the rows in the row block rank1UpdateMultL_Row(blockLength,A,i,i+1,gammasV[A.row0+i]); System.out.println("After update row"); A.original.print(); // apply to the left column block // TODO THIS WON'T WORK!!!!!!!!!!!!! // Needs the whole matrix to have been updated by the left reflector to compute the correct solution // rank1UpdateMultL_LeftCol(blockLength,A,i,i+1,gammasV[A.row0+i]); System.out.println("After row stuff"); A.original.print(); } return true; }
java
@Override public boolean setA(DMatrixRBlock A) { // Extract a lower triangular solution if( !decomposer.decompose(A) ) return false; blockLength = A.blockLength; return true; }
java
@Override public void solve(DMatrixRBlock B, DMatrixRBlock X) { if( B.blockLength != blockLength ) throw new IllegalArgumentException("Unexpected blocklength in B."); DSubmatrixD1 L = new DSubmatrixD1(decomposer.getT(null)); if( X != null ) { if( X.blockLength != blockLength ) throw new IllegalArgumentException("Unexpected blocklength in X."); if( X.numRows != L.col1 ) throw new IllegalArgumentException("Not enough rows in X"); } if( B.numRows != L.col1 ) throw new IllegalArgumentException("Not enough rows in B"); // L * L^T*X = B // Solve for Y: L*Y = B TriangularSolver_DDRB.solve(blockLength,false,L,new DSubmatrixD1(B),false); // L^T * X = Y TriangularSolver_DDRB.solve(blockLength,false,L,new DSubmatrixD1(B),true); if( X != null ) { // copy the solution from B into X MatrixOps_DDRB.extractAligned(B,X); } }
java
public void growInternal(int amount ) { int tmp[] = new int[ data.length + amount ]; System.arraycopy(data,0,tmp,0,data.length); this.data = tmp; }
java
public static boolean lower( double[]T , int indexT , int n ) { double el_ii; double div_el_ii=0; for( int i = 0; i < n; i++ ) { for( int j = i; j < n; j++ ) { double sum = T[ indexT + j*n+i]; // todo optimize for( int k = 0; k < i; k++ ) { sum -= T[ indexT + i*n+k] * T[ indexT + j*n+k]; } if( i == j ) { // is it positive-definite? if( sum <= 0.0 ) return false; el_ii = Math.sqrt(sum); T[ indexT + i*n+i] = el_ii; div_el_ii = 1.0/el_ii; } else { T[ indexT + j*n+i] = sum*div_el_ii; } } } return true; }
java
public static LinearSolverDense<DMatrixRMaj> general(int numRows , int numCols ) { if( numRows == numCols ) return linear(numRows); else return leastSquares(numRows,numCols); }
java
public static LinearSolverDense<DMatrixRMaj> symmPosDef(int matrixWidth ) { if(matrixWidth < EjmlParameters.SWITCH_BLOCK64_CHOLESKY ) { CholeskyDecompositionCommon_DDRM decomp = new CholeskyDecompositionInner_DDRM(true); return new LinearSolverChol_DDRM(decomp); } else { if( EjmlParameters.MEMORY == EjmlParameters.MemoryUsage.FASTER ) return new LinearSolverChol_DDRB(); else { CholeskyDecompositionCommon_DDRM decomp = new CholeskyDecompositionInner_DDRM(true); return new LinearSolverChol_DDRM(decomp); } } }
java
public void setConvergence( int maxIterations , double ftol , double gtol ) { this.maxIterations = maxIterations; this.ftol = ftol; this.gtol = gtol; }
java
private void computeGradientAndHessian(DMatrixRMaj param ) { // residuals = f(x) - y function.compute(param, residuals); computeNumericalJacobian(param,jacobian); CommonOps_DDRM.multTransA(jacobian, residuals, g); CommonOps_DDRM.multTransA(jacobian, jacobian, H); CommonOps_DDRM.extractDiag(H,Hdiag); }
java
public static DMatrixRMaj wrap(int numRows , int numCols , double []data ) { DMatrixRMaj s = new DMatrixRMaj(); s.data = data; s.numRows = numRows; s.numCols = numCols; return s; }
java
public void add( int row , int col , double value ) { if( col < 0 || col >= numCols || row < 0 || row >= numRows ) { throw new IllegalArgumentException("Specified element is out of bounds"); } data[ row * numCols + col ] += value; }
java
@Override public double get( int row , int col ) { if( col < 0 || col >= numCols || row < 0 || row >= numRows ) { throw new IllegalArgumentException("Specified element is out of bounds: "+row+" "+col); } return data[ row * numCols + col ]; }
java
public void set(int numRows, int numCols, boolean rowMajor, double ...data) { reshape(numRows,numCols); int length = numRows*numCols; if( length > this.data.length ) throw new IllegalArgumentException("The length of this matrix's data array is too small."); if( rowMajor ) { System.arraycopy(data,0,this.data,0,length); } else { int index = 0; for( int i = 0; i < numRows; i++ ) { for( int j = 0; j < numCols; j++ ) { this.data[index++] = data[j*numRows+i]; } } } }
java
@Override public void solve(DMatrixRMaj B, DMatrixRMaj X) { blockB.reshape(B.numRows,B.numCols,false); MatrixOps_DDRB.convert(B,blockB); // since overwrite B is true X does not need to be passed in alg.solve(blockB,null); MatrixOps_DDRB.convert(blockB,X); }
java
public List<Complex_F64> getEigenvalues() { List<Complex_F64> ret = new ArrayList<Complex_F64>(); if( is64 ) { EigenDecomposition_F64 d = (EigenDecomposition_F64)eig; for (int i = 0; i < eig.getNumberOfEigenvalues(); i++) { ret.add(d.getEigenvalue(i)); } } else { EigenDecomposition_F32 d = (EigenDecomposition_F32)eig; for (int i = 0; i < eig.getNumberOfEigenvalues(); i++) { Complex_F32 c = d.getEigenvalue(i); ret.add(new Complex_F64(c.real, c.imaginary)); } } return ret; }
java
public int getIndexMax() { int indexMax = 0; double max = getEigenvalue(0).getMagnitude2(); final int N = getNumberOfEigenvalues(); for( int i = 1; i < N; i++ ) { double m = getEigenvalue(i).getMagnitude2(); if( m > max ) { max = m; indexMax = i; } } return indexMax; }
java
public int getIndexMin() { int indexMin = 0; double min = getEigenvalue(0).getMagnitude2(); final int N = getNumberOfEigenvalues(); for( int i = 1; i < N; i++ ) { double m = getEigenvalue(i).getMagnitude2(); if( m < min ) { min = m; indexMin = i; } } return indexMin; }
java
public boolean process( DMatrixSparseCSC A ) { init(A); TriangularSolver_DSCC.eliminationTree(A,true,parent,gwork); countNonZeroInR(parent); countNonZeroInV(parent); // if more columns than rows it's possible that Q*R != A. That's because a householder // would need to be created that's outside the m by m Q matrix. In reality it has // a partial solution. Column pivot are needed. if( m < n ) { for (int row = 0; row <m; row++) { if( gwork.data[head+row] < 0 ) { return false; } } } return true; }
java
void init( DMatrixSparseCSC A ) { this.A = A; this.m = A.numRows; this.n = A.numCols; this.next = 0; this.head = m; this.tail = m + n; this.nque = m + 2*n; if( parent.length < n || leftmost.length < m) { parent = new int[n]; post = new int[n]; pinv = new int[m+n]; countsR = new int[n]; leftmost = new int[m]; } gwork.reshape(m+3*n); }
java
void countNonZeroInR( int[] parent ) { TriangularSolver_DSCC.postorder(parent,n,post,gwork); columnCounts.process(A,parent,post,countsR); nz_in_R = 0; for (int k = 0; k < n; k++) { nz_in_R += countsR[k]; } if( nz_in_R < 0) throw new RuntimeException("Too many elements. Numerical overflow in R counts"); }
java
void countNonZeroInV( int []parent ) { int []w = gwork.data; findMinElementIndexInRows(leftmost); createRowElementLinkedLists(leftmost,w); countNonZeroUsingLinkedList(parent,w); }
java
void countNonZeroUsingLinkedList( int parent[] , int ll[] ) { Arrays.fill(pinv,0,m,-1); nz_in_V = 0; m2 = m; for (int k = 0; k < n; k++) { int i = ll[head+k]; // remove row i from queue k nz_in_V++; // count V(k,k) as nonzero if( i < 0) // add a fictitious row since there are no nz elements i = m2++; pinv[i] = k; // associate row i with V(:,k) if( --ll[nque+k] <= 0 ) continue; nz_in_V += ll[nque+k]; int pa; if( (pa = parent[k]) != -1 ) { // move all rows to parent of k if( ll[nque+pa] == 0) ll[tail+pa] = ll[tail+k]; ll[next+ll[tail+k]] = ll[head+pa]; ll[head+pa] = ll[next+i]; ll[nque+pa] += ll[nque+k]; } } for (int i = 0, k = n; i < m; i++) { if( pinv[i] < 0 ) pinv[i] = k++; } if( nz_in_V < 0) throw new RuntimeException("Too many elements. Numerical overflow in V counts"); }
java
public void alias(DMatrixRMaj variable , String name ) { if( isReserved(name)) throw new RuntimeException("Reserved word or contains a reserved character"); VariableMatrix old = (VariableMatrix)variables.get(name); if( old == null ) { variables.put(name, new VariableMatrix(variable)); }else { old.matrix = variable; } }
java
public void alias( double value , String name ) { if( isReserved(name)) throw new RuntimeException("Reserved word or contains a reserved character. '"+name+"'"); VariableDouble old = (VariableDouble)variables.get(name); if( old == null ) { variables.put(name, new VariableDouble(value)); }else { old.value = value; } }
java
public void alias( Object ...args ) { if( args.length % 2 == 1 ) throw new RuntimeException("Even number of arguments expected"); for (int i = 0; i < args.length; i += 2) { aliasGeneric( args[i], (String)args[i+1]); } }
java
protected void aliasGeneric( Object variable , String name ) { if( variable.getClass() == Integer.class ) { alias(((Integer)variable).intValue(),name); } else if( variable.getClass() == Double.class ) { alias(((Double)variable).doubleValue(),name); } else if( variable.getClass() == DMatrixRMaj.class ) { alias((DMatrixRMaj)variable,name); } else if( variable.getClass() == FMatrixRMaj.class ) { alias((FMatrixRMaj)variable,name); } else if( variable.getClass() == DMatrixSparseCSC.class ) { alias((DMatrixSparseCSC)variable,name); } else if( variable.getClass() == SimpleMatrix.class ) { alias((SimpleMatrix) variable, name); } else if( variable instanceof DMatrixFixed ) { DMatrixRMaj M = new DMatrixRMaj(1,1); ConvertDMatrixStruct.convert((DMatrixFixed)variable,M); alias(M,name); } else if( variable instanceof FMatrixFixed ) { FMatrixRMaj M = new FMatrixRMaj(1,1); ConvertFMatrixStruct.convert((FMatrixFixed)variable,M); alias(M,name); } else { throw new RuntimeException("Unknown value type of "+ (variable.getClass().getSimpleName())+" for variable "+name); } }
java
public Sequence compile( String equation , boolean assignment, boolean debug ) { functions.setManagerTemp(managerTemp); Sequence sequence = new Sequence(); TokenList tokens = extractTokens(equation,managerTemp); if( tokens.size() < 3 ) throw new RuntimeException("Too few tokens"); TokenList.Token t0 = tokens.getFirst(); if( t0.word != null && t0.word.compareToIgnoreCase("macro") == 0 ) { parseMacro(tokens,sequence); } else { insertFunctionsAndVariables(tokens); insertMacros(tokens); if (debug) { System.out.println("Parsed tokens:\n------------"); tokens.print(); System.out.println(); } // Get the results variable if (t0.getType() != Type.VARIABLE && t0.getType() != Type.WORD) { compileTokens(sequence,tokens); // If there's no output then this is acceptable, otherwise it's assumed to be a bug // If there's no output then a configuration was changed Variable variable = tokens.getFirst().getVariable(); if( variable != null ) { if( assignment ) throw new IllegalArgumentException("No assignment to an output variable could be found. Found " + t0); else { sequence.output = variable; // set this to be the output for print() } } } else { compileAssignment(sequence, tokens, t0); } if (debug) { System.out.println("Operations:\n------------"); for (int i = 0; i < sequence.operations.size(); i++) { System.out.println(sequence.operations.get(i).name()); } } } return sequence; }
java
private void parseMacro( TokenList tokens , Sequence sequence ) { Macro macro = new Macro(); TokenList.Token t = tokens.getFirst().next; if( t.word == null ) { throw new ParseError("Expected the macro's name after "+tokens.getFirst().word); } List<TokenList.Token> variableTokens = new ArrayList<TokenList.Token>(); macro.name = t.word; t = t.next; t = parseMacroInput(variableTokens, t); for( TokenList.Token a : variableTokens ) { if( a.word == null) throw new ParseError("expected word in macro header"); macro.inputs.add(a.word); } t = t.next; if( t == null || t.getSymbol() != Symbol.ASSIGN) throw new ParseError("Expected assignment"); t = t.next; macro.tokens = new TokenList(t,tokens.last); sequence.addOperation(macro.createOperation(macros)); }
java
private void checkForUnknownVariables(TokenList tokens) { TokenList.Token t = tokens.getFirst(); while( t != null ) { if( t.getType() == Type.WORD ) throw new ParseError("Unknown variable on right side. "+t.getWord()); t = t.next; } }
java
private Variable createVariableInferred(TokenList.Token t0, Variable variableRight) { Variable result; if( t0.getType() == Type.WORD ) { switch( variableRight.getType()) { case MATRIX: alias(new DMatrixRMaj(1,1),t0.getWord()); break; case SCALAR: if( variableRight instanceof VariableInteger) { alias(0,t0.getWord()); } else { alias(1.0,t0.getWord()); } break; case INTEGER_SEQUENCE: alias((IntegerSequence)null,t0.getWord()); break; default: throw new RuntimeException("Type not supported for assignment: "+variableRight.getType()); } result = variables.get(t0.getWord()); } else { result = t0.getVariable(); } return result; }
java
private List<Variable> parseAssignRange(Sequence sequence, TokenList tokens, TokenList.Token t0) { // find assignment symbol TokenList.Token tokenAssign = t0.next; while( tokenAssign != null && tokenAssign.symbol != Symbol.ASSIGN ) { tokenAssign = tokenAssign.next; } if( tokenAssign == null ) throw new ParseError("Can't find assignment operator"); // see if it is a sub matrix before if( tokenAssign.previous.symbol == Symbol.PAREN_RIGHT ) { TokenList.Token start = t0.next; if( start.symbol != Symbol.PAREN_LEFT ) throw new ParseError(("Expected left param for assignment")); TokenList.Token end = tokenAssign.previous; TokenList subTokens = tokens.extractSubList(start,end); subTokens.remove(subTokens.getFirst()); subTokens.remove(subTokens.getLast()); handleParentheses(subTokens,sequence); List<TokenList.Token> inputs = parseParameterCommaBlock(subTokens, sequence); if (inputs.isEmpty()) throw new ParseError("Empty function input parameters"); List<Variable> range = new ArrayList<>(); addSubMatrixVariables(inputs, range); if( range.size() != 1 && range.size() != 2 ) { throw new ParseError("Unexpected number of range variables. 1 or 2 expected"); } return range; } return null; }
java
protected void handleParentheses( TokenList tokens, Sequence sequence ) { // have a list to handle embedded parentheses, e.g. (((((a))))) List<TokenList.Token> left = new ArrayList<TokenList.Token>(); // find all of them TokenList.Token t = tokens.first; while( t != null ) { TokenList.Token next = t.next; if( t.getType() == Type.SYMBOL ) { if( t.getSymbol() == Symbol.PAREN_LEFT ) left.add(t); else if( t.getSymbol() == Symbol.PAREN_RIGHT ) { if( left.isEmpty() ) throw new ParseError(") found with no matching ("); TokenList.Token a = left.remove(left.size()-1); // remember the element before so the new one can be inserted afterwards TokenList.Token before = a.previous; TokenList sublist = tokens.extractSubList(a,t); // remove parentheses sublist.remove(sublist.first); sublist.remove(sublist.last); // if its a function before () then the () indicates its an input to a function if( before != null && before.getType() == Type.FUNCTION ) { List<TokenList.Token> inputs = parseParameterCommaBlock(sublist, sequence); if (inputs.isEmpty()) throw new ParseError("Empty function input parameters"); else { createFunction(before, inputs, tokens, sequence); } } else if( before != null && before.getType() == Type.VARIABLE && before.getVariable().getType() == VariableType.MATRIX ) { // if it's a variable then that says it's a sub-matrix TokenList.Token extract = parseSubmatrixToExtract(before,sublist, sequence); // put in the extract operation tokens.insert(before,extract); tokens.remove(before); } else { // if null then it was empty inside TokenList.Token output = parseBlockNoParentheses(sublist,sequence, false); if (output != null) tokens.insert(before, output); } } } t = next; } if( !left.isEmpty()) throw new ParseError("Dangling ( parentheses"); }
java
protected List<TokenList.Token> parseParameterCommaBlock( TokenList tokens, Sequence sequence ) { // find all the comma tokens List<TokenList.Token> commas = new ArrayList<TokenList.Token>(); TokenList.Token token = tokens.first; int numBracket = 0; while( token != null ) { if( token.getType() == Type.SYMBOL ) { switch( token.getSymbol() ) { case COMMA: if( numBracket == 0) commas.add(token); break; case BRACKET_LEFT: numBracket++; break; case BRACKET_RIGHT: numBracket--; break; } } token = token.next; } List<TokenList.Token> output = new ArrayList<TokenList.Token>(); if( commas.isEmpty() ) { output.add(parseBlockNoParentheses(tokens, sequence, false)); } else { TokenList.Token before = tokens.first; for (int i = 0; i < commas.size(); i++) { TokenList.Token after = commas.get(i); if( before == after ) throw new ParseError("No empty function inputs allowed!"); TokenList.Token tmp = after.next; TokenList sublist = tokens.extractSubList(before,after); sublist.remove(after);// remove the comma output.add(parseBlockNoParentheses(sublist, sequence, false)); before = tmp; } // if the last character is a comma then after.next above will be null and thus before is null if( before == null ) throw new ParseError("No empty function inputs allowed!"); TokenList.Token after = tokens.last; TokenList sublist = tokens.extractSubList(before, after); output.add(parseBlockNoParentheses(sublist, sequence, false)); } return output; }
java
protected TokenList.Token parseSubmatrixToExtract(TokenList.Token variableTarget, TokenList tokens, Sequence sequence) { List<TokenList.Token> inputs = parseParameterCommaBlock(tokens, sequence); List<Variable> variables = new ArrayList<Variable>(); // for the operation, the first variable must be the matrix which is being manipulated variables.add(variableTarget.getVariable()); addSubMatrixVariables(inputs, variables); if( variables.size() != 2 && variables.size() != 3 ) { throw new ParseError("Unexpected number of variables. 1 or 2 expected"); } // first parameter is the matrix it will be extracted from. rest specify range Operation.Info info; // only one variable means its referencing elements // two variables means its referencing a sub matrix if( inputs.size() == 1 ) { Variable varA = variables.get(1); if( varA.getType() == VariableType.SCALAR ) { info = functions.create("extractScalar", variables); } else { info = functions.create("extract", variables); } } else if( inputs.size() == 2 ) { Variable varA = variables.get(1); Variable varB = variables.get(2); if( varA.getType() == VariableType.SCALAR && varB.getType() == VariableType.SCALAR) { info = functions.create("extractScalar", variables); } else { info = functions.create("extract", variables); } } else { throw new ParseError("Expected 2 inputs to sub-matrix"); } sequence.addOperation(info.op); return new TokenList.Token(info.output); }
java
private void addSubMatrixVariables(List<TokenList.Token> inputs, List<Variable> variables) { for (int i = 0; i < inputs.size(); i++) { TokenList.Token t = inputs.get(i); if( t.getType() != Type.VARIABLE ) throw new ParseError("Expected variables only in sub-matrix input, not "+t.getType()); Variable v = t.getVariable(); if( v.getType() == VariableType.INTEGER_SEQUENCE || isVariableInteger(t) ) { variables.add(v); } else { throw new ParseError("Expected an integer, integer sequence, or array range to define a submatrix"); } } }
java
protected TokenList.Token parseBlockNoParentheses(TokenList tokens, Sequence sequence, boolean insideMatrixConstructor) { // search for matrix bracket operations if( !insideMatrixConstructor ) { parseBracketCreateMatrix(tokens, sequence); } // First create sequences from anything involving a colon parseSequencesWithColons(tokens, sequence ); // process operators depending on their priority parseNegOp(tokens, sequence); parseOperationsL(tokens, sequence); parseOperationsLR(new Symbol[]{Symbol.POWER, Symbol.ELEMENT_POWER}, tokens, sequence); parseOperationsLR(new Symbol[]{Symbol.TIMES, Symbol.RDIVIDE, Symbol.LDIVIDE, Symbol.ELEMENT_TIMES, Symbol.ELEMENT_DIVIDE}, tokens, sequence); parseOperationsLR(new Symbol[]{Symbol.PLUS, Symbol.MINUS}, tokens, sequence); // Commas are used in integer sequences. Can be used to force to compiler to treat - as negative not // minus. They can now be removed since they have served their purpose stripCommas(tokens); // now construct rest of the lists and combine them together parseIntegerLists(tokens); parseCombineIntegerLists(tokens); if( !insideMatrixConstructor ) { if (tokens.size() > 1) { System.err.println("Remaining tokens: "+tokens.size); TokenList.Token t = tokens.first; while( t != null ) { System.err.println(" "+t); t = t.next; } throw new RuntimeException("BUG in parser. There should only be a single token left"); } return tokens.first; } else { return null; } }
java
private void stripCommas(TokenList tokens) { TokenList.Token t = tokens.getFirst(); while( t != null ) { TokenList.Token next = t.next; if( t.getSymbol() == Symbol.COMMA ) { tokens.remove(t); } t = next; } }
java
protected void parseSequencesWithColons(TokenList tokens , Sequence sequence ) { TokenList.Token t = tokens.getFirst(); if( t == null ) return; int state = 0; TokenList.Token start = null; TokenList.Token middle = null; TokenList.Token prev = t; boolean last = false; while( true ) { if( state == 0 ) { if( isVariableInteger(t) && (t.next != null && t.next.getSymbol() == Symbol.COLON) ) { start = t; state = 1; t = t.next; } else if( t != null && t.getSymbol() == Symbol.COLON ) { // If it starts with a colon then it must be 'all' or a type-o IntegerSequence range = new IntegerSequence.Range(null,null); VariableIntegerSequence varSequence = functions.getManagerTemp().createIntegerSequence(range); TokenList.Token n = new TokenList.Token(varSequence); tokens.insert(t.previous, n); tokens.remove(t); t = n; } } else if( state == 1 ) { // var : ? if (isVariableInteger(t)) { state = 2; } else { // array range IntegerSequence range = new IntegerSequence.Range(start,null); VariableIntegerSequence varSequence = functions.getManagerTemp().createIntegerSequence(range); replaceSequence(tokens, varSequence, start, prev); state = 0; } } else if ( state == 2 ) { // var:var ? if( t != null && t.getSymbol() == Symbol.COLON ) { middle = prev; state = 3; } else { // create for sequence with start and stop elements only IntegerSequence numbers = new IntegerSequence.For(start,null,prev); VariableIntegerSequence varSequence = functions.getManagerTemp().createIntegerSequence(numbers); replaceSequence(tokens, varSequence, start, prev ); if( t != null ) t = t.previous; state = 0; } } else if ( state == 3 ) { // var:var: ? if( isVariableInteger(t) ) { // create 'for' sequence with three variables IntegerSequence numbers = new IntegerSequence.For(start,middle,t); VariableIntegerSequence varSequence = functions.getManagerTemp().createIntegerSequence(numbers); t = replaceSequence(tokens, varSequence, start, t); } else { // array range with 2 elements IntegerSequence numbers = new IntegerSequence.Range(start,middle); VariableIntegerSequence varSequence = functions.getManagerTemp().createIntegerSequence(numbers); replaceSequence(tokens, varSequence, start, prev); } state = 0; } if( last ) { break; } else if( t.next == null ) { // handle the case where it is the last token in the sequence last = true; } prev = t; t = t.next; } }
java
protected void parseIntegerLists(TokenList tokens) { TokenList.Token t = tokens.getFirst(); if( t == null || t.next == null ) return; int state = 0; TokenList.Token start = null; TokenList.Token prev = t; boolean last = false; while( true ) { if( state == 0 ) { if( isVariableInteger(t) ) { start = t; state = 1; } } else if( state == 1 ) { // var ? if( isVariableInteger(t)) { // see if its explicit number sequence state = 2; } else { // just scalar integer, skip state = 0; } } else if ( state == 2 ) { // var var .... if( !isVariableInteger(t) ) { // create explicit list sequence IntegerSequence sequence = new IntegerSequence.Explicit(start,prev); VariableIntegerSequence varSequence = functions.getManagerTemp().createIntegerSequence(sequence); replaceSequence(tokens, varSequence, start, prev); state = 0; } } if( last ) { break; } else if( t.next == null ) { // handle the case where it is the last token in the sequence last = true; } prev = t; t = t.next; } }
java
protected void parseCombineIntegerLists(TokenList tokens) { TokenList.Token t = tokens.getFirst(); if( t == null || t.next == null ) return; int numFound = 0; TokenList.Token start = null; TokenList.Token end = null; while( t != null ) { if( t.getType() == Type.VARIABLE && (isVariableInteger(t) || t.getVariable().getType() == VariableType.INTEGER_SEQUENCE )) { if( numFound == 0 ) { numFound = 1; start = end = t; } else { numFound++; end = t; } } else if( numFound > 1 ) { IntegerSequence sequence = new IntegerSequence.Combined(start,end); VariableIntegerSequence varSequence = functions.getManagerTemp().createIntegerSequence(sequence); replaceSequence(tokens, varSequence, start, end); numFound = 0; } else { numFound = 0; } t = t.next; } if( numFound > 1 ) { IntegerSequence sequence = new IntegerSequence.Combined(start,end); VariableIntegerSequence varSequence = functions.getManagerTemp().createIntegerSequence(sequence); replaceSequence(tokens, varSequence, start, end); } }
java
private static boolean isVariableInteger(TokenList.Token t) { if( t == null ) return false; return t.getScalarType() == VariableScalar.Type.INTEGER; }
java
protected void parseBracketCreateMatrix(TokenList tokens, Sequence sequence) { List<TokenList.Token> left = new ArrayList<TokenList.Token>(); TokenList.Token t = tokens.getFirst(); while( t != null ) { TokenList.Token next = t.next; if( t.getSymbol() == Symbol.BRACKET_LEFT ) { left.add(t); } else if( t.getSymbol() == Symbol.BRACKET_RIGHT ) { if( left.isEmpty() ) throw new RuntimeException("No matching left bracket for right"); TokenList.Token start = left.remove(left.size() - 1); // Compute everything inside the [ ], this will leave a // series of variables and semi-colons hopefully TokenList bracketLet = tokens.extractSubList(start.next,t.previous); parseBlockNoParentheses(bracketLet, sequence, true); MatrixConstructor constructor = constructMatrix(bracketLet); // define the matrix op and inject into token list Operation.Info info = Operation.matrixConstructor(constructor); sequence.addOperation(info.op); tokens.insert(start.previous, new TokenList.Token(info.output)); // remove the brackets tokens.remove(start); tokens.remove(t); } t = next; } if( !left.isEmpty() ) throw new RuntimeException("Dangling ["); }
java
protected void parseNegOp(TokenList tokens, Sequence sequence) { if( tokens.size == 0 ) return; TokenList.Token token = tokens.first; while( token != null ) { TokenList.Token next = token.next; escape: if( token.getSymbol() == Symbol.MINUS ) { if( token.previous != null && token.previous.getType() != Type.SYMBOL) break escape; if( token.previous != null && token.previous.getType() == Type.SYMBOL && (token.previous.symbol == Symbol.TRANSPOSE)) break escape; if( token.next == null || token.next.getType() == Type.SYMBOL) break escape; if( token.next.getType() != Type.VARIABLE ) throw new RuntimeException("Crap bug rethink this function"); // create the operation Operation.Info info = Operation.neg(token.next.getVariable(),functions.getManagerTemp()); // add the operation to the sequence sequence.addOperation(info.op); // update the token list TokenList.Token t = new TokenList.Token(info.output); tokens.insert(token.next,t); tokens.remove(token.next); tokens.remove(token); next = t; } token = next; } }
java
protected void parseOperationsL(TokenList tokens, Sequence sequence) { if( tokens.size == 0 ) return; TokenList.Token token = tokens.first; if( token.getType() != Type.VARIABLE ) throw new ParseError("The first token in an equation needs to be a variable and not "+token); while( token != null ) { if( token.getType() == Type.FUNCTION ) { throw new ParseError("Function encountered with no parentheses"); } else if( token.getType() == Type.SYMBOL && token.getSymbol() == Symbol.TRANSPOSE) { if( token.previous.getType() == Type.VARIABLE ) token = insertTranspose(token.previous,tokens,sequence); else throw new ParseError("Expected variable before transpose"); } token = token.next; } }
java
protected void parseOperationsLR(Symbol ops[], TokenList tokens, Sequence sequence) { if( tokens.size == 0 ) return; TokenList.Token token = tokens.first; if( token.getType() != Type.VARIABLE ) throw new ParseError("The first token in an equation needs to be a variable and not "+token); boolean hasLeft = false; while( token != null ) { if( token.getType() == Type.FUNCTION ) { throw new ParseError("Function encountered with no parentheses"); } else if( token.getType() == Type.VARIABLE ) { if( hasLeft ) { if( isTargetOp(token.previous,ops)) { token = createOp(token.previous.previous,token.previous,token,tokens,sequence); } } else { hasLeft = true; } } else { if( token.previous.getType() == Type.SYMBOL ) { throw new ParseError("Two symbols next to each other. "+token.previous+" and "+token); } } token = token.next; } }
java
public <T extends Variable> T lookupVariable(String token) { Variable result = variables.get(token); return (T)result; }
java
void insertMacros(TokenList tokens ) { TokenList.Token t = tokens.getFirst(); while( t != null ) { if( t.getType() == Type.WORD ) { Macro v = lookupMacro(t.word); if (v != null) { TokenList.Token before = t.previous; List<TokenList.Token> inputs = new ArrayList<TokenList.Token>(); t = parseMacroInput(inputs,t.next); TokenList sniplet = v.execute(inputs); tokens.extractSubList(before.next,t); tokens.insertAfter(before,sniplet); t = sniplet.last; } } t = t.next; } }
java
protected static boolean isTargetOp( TokenList.Token token , Symbol[] ops ) { Symbol c = token.symbol; for (int i = 0; i < ops.length; i++) { if( c == ops[i]) return true; } return false; }
java
protected static boolean isOperatorLR( Symbol s ) { if( s == null ) return false; switch( s ) { case ELEMENT_DIVIDE: case ELEMENT_TIMES: case ELEMENT_POWER: case RDIVIDE: case LDIVIDE: case TIMES: case POWER: case PLUS: case MINUS: case ASSIGN: return true; } return false; }
java
protected boolean isReserved( String name ) { if( functions.isFunctionName(name)) return true; for (int i = 0; i < name.length(); i++) { if( !isLetter(name.charAt(i)) ) return true; } return false; }
java
public Equation process( String equation , boolean debug ) { compile(equation,true,debug).perform(); return this; }
java
public void print( String equation ) { // first assume it's just a variable Variable v = lookupVariable(equation); if( v == null ) { Sequence sequence = compile(equation,false,false); sequence.perform(); v = sequence.output; } if( v instanceof VariableMatrix ) { ((VariableMatrix)v).matrix.print(); } else if(v instanceof VariableScalar ) { System.out.println("Scalar = "+((VariableScalar)v).getDouble() ); } else { System.out.println("Add support for "+v.getClass().getSimpleName()); } }
java
public static double computeTauAndDivide(final int j, final int numRows , final double[] u , final double max) { double tau = 0; // double div_max = 1.0/max; // if( Double.isInfinite(div_max)) { for( int i = j; i < numRows; i++ ) { double d = u[i] /= max; tau += d*d; } // } else { // for( int i = j; i < numRows; i++ ) { // double d = u[i] *= div_max; // tau += d*d; // } // } tau = Math.sqrt(tau); if( u[j] < 0 ) tau = -tau; return tau; }
java
public static boolean isSameStructure(DMatrixSparseCSC a , DMatrixSparseCSC b) { if( a.numRows == b.numRows && a.numCols == b.numCols && a.nz_length == b.nz_length) { for (int i = 0; i <= a.numCols; i++) { if( a.col_idx[i] != b.col_idx[i] ) return false; } for (int i = 0; i < a.nz_length; i++) { if( a.nz_rows[i] != b.nz_rows[i] ) return false; } return true; } return false; }
java
public static boolean isVector(DMatrixSparseCSC a) { return (a.numCols == 1 && a.numRows > 1) || (a.numRows == 1 && a.numCols>1); }
java
public static boolean isSymmetric( DMatrixSparseCSC A , double tol ) { if( A.numRows != A.numCols ) return false; int N = A.numCols; for (int i = 0; i < N; i++) { int idx0 = A.col_idx[i]; int idx1 = A.col_idx[i+1]; for (int index = idx0; index < idx1; index++) { int j = A.nz_rows[index]; double value_ji = A.nz_values[index]; double value_ij = A.get(i,j); if( Math.abs(value_ij-value_ji) > tol ) return false; } } return true; }
java
public void implicitDoubleStep( int x1 , int x2 ) { if( printHumps ) System.out.println("Performing implicit double step"); // compute the wilkinson shift double z11 = A.get(x2 - 1, x2 - 1); double z12 = A.get(x2 - 1, x2); double z21 = A.get(x2, x2 - 1); double z22 = A.get(x2, x2); double a11 = A.get(x1,x1); double a21 = A.get(x1+1,x1); double a12 = A.get(x1,x1+1); double a22 = A.get(x1+1,x1+1); double a32 = A.get(x1+2,x1+1); if( normalize ) { temp[0] = a11;temp[1] = a21;temp[2] = a12;temp[3] = a22;temp[4] = a32; temp[5] = z11;temp[6] = z22;temp[7] = z12;temp[8] = z21; double max = Math.abs(temp[0]); for( int j = 1; j < temp.length; j++ ) { if( Math.abs(temp[j]) > max ) max = Math.abs(temp[j]); } a11 /= max;a21 /= max;a12 /= max;a22 /= max;a32 /= max; z11 /= max;z22 /= max;z12 /= max;z21 /= max; } // these equations are derived when the eigenvalues are extracted from the lower right // 2 by 2 matrix. See page 388 of Fundamentals of Matrix Computations 2nd ed for details. double b11,b21,b31; if( useStandardEq ) { b11 = ((a11- z11)*(a11- z22)- z21 * z12)/a21 + a12; b21 = a11 + a22 - z11 - z22; b31 = a32; } else { // this is different from the version in the book and seems in my testing to be more resilient to // over flow issues b11 = ((a11- z11)*(a11- z22)- z21 * z12) + a12*a21; b21 = (a11 + a22 - z11 - z22)*a21; b31 = a32*a21; } performImplicitDoubleStep(x1, x2, b11 , b21 , b31 ); }
java
public void performImplicitDoubleStep(int x1, int x2 , double real , double img ) { double a11 = A.get(x1,x1); double a21 = A.get(x1+1,x1); double a12 = A.get(x1,x1+1); double a22 = A.get(x1+1,x1+1); double a32 = A.get(x1+2,x1+1); double p_plus_t = 2.0*real; double p_times_t = real*real + img*img; double b11,b21,b31; if( useStandardEq ) { b11 = (a11*a11 - p_plus_t*a11+p_times_t)/a21 + a12; b21 = a11+a22-p_plus_t; b31 = a32; } else { // this is different from the version in the book and seems in my testing to be more resilient to // over flow issues b11 = (a11*a11 - p_plus_t*a11+p_times_t) + a12*a21; b21 = (a11+a22-p_plus_t)*a21; b31 = a32*a21; } performImplicitDoubleStep(x1, x2, b11, b21, b31); }
java
public boolean process(DMatrixRMaj A , int numSingularValues, DMatrixRMaj nullspace ) { decomposition.decompose(A); if( A.numRows > A.numCols ) { Q.reshape(A.numCols,Math.min(A.numRows,A.numCols)); decomposition.getQ(Q, true); } else { Q.reshape(A.numCols, A.numCols); decomposition.getQ(Q, false); } nullspace.reshape(Q.numRows,numSingularValues); CommonOps_DDRM.extract(Q,0,Q.numRows,Q.numCols-numSingularValues,Q.numCols,nullspace,0,0); return true; }
java
public static boolean isInverse(DMatrixRMaj a , DMatrixRMaj b , double tol ) { if( a.numRows != b.numRows || a.numCols != b.numCols ) { return false; } int numRows = a.numRows; int numCols = a.numCols; for( int i = 0; i < numRows; i++ ) { for( int j = 0; j < numCols; j++ ) { double total = 0; for( int k = 0; k < numCols; k++ ) { total += a.get(i,k)*b.get(k,j); } if( i == j ) { if( !(Math.abs(total-1) <= tol) ) return false; } else if( !(Math.abs(total) <= tol) ) return false; } } return true; }
java
public static boolean isRowsLinearIndependent( DMatrixRMaj A ) { // LU decomposition LUDecomposition<DMatrixRMaj> lu = DecompositionFactory_DDRM.lu(A.numRows,A.numCols); if( lu.inputModified() ) A = A.copy(); if( !lu.decompose(A)) throw new RuntimeException("Decompositon failed?"); // if they are linearly independent it should not be singular return !lu.isSingular(); }
java
public static boolean isConstantVal(DMatrixRMaj mat , double val , double tol ) { // see if the result is an identity matrix int index = 0; for( int i = 0; i < mat.numRows; i++ ) { for( int j = 0; j < mat.numCols; j++ ) { if( !(Math.abs(mat.get(index++)-val) <= tol) ) return false; } } return true; }
java
public static boolean isDiagonalPositive( DMatrixRMaj a ) { for( int i = 0; i < a.numRows; i++ ) { if( !(a.get(i,i) >= 0) ) return false; } return true; }
java
public static int rank(DMatrixRMaj A , double threshold ) { SingularValueDecomposition_F64<DMatrixRMaj> svd = DecompositionFactory_DDRM.svd(A.numRows,A.numCols,false,false,true); if( svd.inputModified() ) A = A.copy(); if( !svd.decompose(A) ) throw new RuntimeException("Decomposition failed"); return SingularOps_DDRM.rank(svd, threshold); }
java
public static int countNonZero(DMatrixRMaj A){ int total = 0; for (int row = 0, index=0; row < A.numRows; row++) { for (int col = 0; col < A.numCols; col++,index++) { if( A.data[index] != 0 ) { total++; } } } return total; }
java
public static boolean invertSPD(DMatrixRMaj mat, DMatrixRMaj result ) { if( mat.numRows != mat.numCols ) throw new IllegalArgumentException("Must be a square matrix"); result.reshape(mat.numRows,mat.numRows); if( mat.numRows <= UnrolledCholesky_DDRM.MAX ) { // L*L' = A if( !UnrolledCholesky_DDRM.lower(mat,result) ) return false; // L = inv(L) TriangularSolver_DDRM.invertLower(result.data,result.numCols); // inv(A) = inv(L')*inv(L) SpecializedOps_DDRM.multLowerTranA(result); } else { LinearSolverDense<DMatrixRMaj> solver = LinearSolverFactory_DDRM.chol(mat.numCols); if( solver.modifiesA() ) mat = mat.copy(); if( !solver.setA(mat)) return false; solver.invert(result); } return true; }
java
public static DMatrixRMaj identity(int numRows , int numCols ) { DMatrixRMaj ret = new DMatrixRMaj(numRows,numCols); int small = numRows < numCols ? numRows : numCols; for( int i = 0; i < small; i++ ) { ret.set(i,i,1.0); } return ret; }
java
public static void extract( DMatrix src, int srcY0, int srcY1, int srcX0, int srcX1, DMatrix dst ) { ((ReshapeMatrix)dst).reshape(srcY1-srcY0,srcX1-srcX0); extract(src,srcY0,srcY1,srcX0,srcX1,dst,0,0); }
java
public static void extract( DMatrixRMaj src, int rows[] , int rowsSize , int cols[] , int colsSize , DMatrixRMaj dst ) { if( rowsSize != dst.numRows || colsSize != dst.numCols ) throw new MatrixDimensionException("Unexpected number of rows and/or columns in dst matrix"); int indexDst = 0; for (int i = 0; i < rowsSize; i++) { int indexSrcRow = src.numCols*rows[i]; for (int j = 0; j < colsSize; j++) { dst.data[indexDst++] = src.data[indexSrcRow + cols[j]]; } } }
java
public static void extract(DMatrixRMaj src, int indexes[] , int length , DMatrixRMaj dst ) { if( !MatrixFeatures_DDRM.isVector(dst)) throw new MatrixDimensionException("Dst must be a vector"); if( length != dst.getNumElements()) throw new MatrixDimensionException("Unexpected number of elements in dst vector"); for (int i = 0; i < length; i++) { dst.data[i] = src.data[indexes[i]]; } }
java
public static DMatrixRMaj extractRow(DMatrixRMaj a , int row , DMatrixRMaj out ) { if( out == null) out = new DMatrixRMaj(1,a.numCols); else if( !MatrixFeatures_DDRM.isVector(out) || out.getNumElements() != a.numCols ) throw new MatrixDimensionException("Output must be a vector of length "+a.numCols); System.arraycopy(a.data,a.getIndex(row,0),out.data,0,a.numCols); return out; }
java
public static DMatrixRMaj extractColumn(DMatrixRMaj a , int column , DMatrixRMaj out ) { if( out == null) out = new DMatrixRMaj(a.numRows,1); else if( !MatrixFeatures_DDRM.isVector(out) || out.getNumElements() != a.numRows ) throw new MatrixDimensionException("Output must be a vector of length "+a.numRows); int index = column; for (int i = 0; i < a.numRows; i++, index += a.numCols ) { out.data[i] = a.data[index]; } return out; }
java
public static void removeColumns( DMatrixRMaj A , int col0 , int col1 ) { if( col1 < col0 ) { throw new IllegalArgumentException("col1 must be >= col0"); } else if( col0 >= A.numCols || col1 >= A.numCols ) { throw new IllegalArgumentException("Columns which are to be removed must be in bounds"); } int step = col1-col0+1; int offset = 0; for (int row = 0, idx=0; row < A.numRows; row++) { for (int i = 0; i < col0; i++,idx++) { A.data[idx] = A.data[idx+offset]; } offset += step; for (int i = col1+1; i < A.numCols; i++,idx++) { A.data[idx] = A.data[idx+offset]; } } A.numCols -= step; }
java
public static void scaleRow( double alpha , DMatrixRMaj A , int row ) { int idx = row*A.numCols; for (int col = 0; col < A.numCols; col++) { A.data[idx++] *= alpha; } }
java
public static void scaleCol( double alpha , DMatrixRMaj A , int col ) { int idx = col; for (int row = 0; row < A.numRows; row++, idx += A.numCols) { A.data[idx] *= alpha; } }
java
public static BMatrixRMaj elementLessThan(DMatrixRMaj A , double value , BMatrixRMaj output ) { if( output == null ) { output = new BMatrixRMaj(A.numRows,A.numCols); } output.reshape(A.numRows, A.numCols); int N = A.getNumElements(); for (int i = 0; i < N; i++) { output.data[i] = A.data[i] < value; } return output; }
java
public static DMatrixRMaj elements(DMatrixRMaj A , BMatrixRMaj marked , DMatrixRMaj output ) { if( A.numRows != marked.numRows || A.numCols != marked.numCols ) throw new MatrixDimensionException("Input matrices must have the same shape"); if( output == null ) output = new DMatrixRMaj(1,1); output.reshape(countTrue(marked),1); int N = A.getNumElements(); int index = 0; for (int i = 0; i < N; i++) { if( marked.data[i] ) { output.data[index++] = A.data[i]; } } return output; }
java
public static int countTrue(BMatrixRMaj A) { int total = 0; int N = A.getNumElements(); for (int i = 0; i < N; i++) { if( A.data[i] ) total++; } return total; }
java
public static void symmLowerToFull( DMatrixRMaj A ) { if( A.numRows != A.numCols ) throw new MatrixDimensionException("Must be a square matrix"); final int cols = A.numCols; for (int row = 0; row < A.numRows; row++) { for (int col = row+1; col < cols; col++) { A.data[row*cols+col] = A.data[col*cols+row]; } } }
java
public void init( DMatrixRMaj A ) { if( A.numRows != A.numCols) throw new IllegalArgumentException("Must be square"); if( A.numCols != N ) { N = A.numCols; QT.reshape(N,N, false); if( w.length < N ) { w = new double[ N ]; gammas = new double[N]; b = new double[N]; } } // just copy the top right triangle QT.set(A); }
java
public static void inner_reorder_lower(DMatrix1Row A , DMatrix1Row B ) { final int cols = A.numCols; B.reshape(cols,cols); Arrays.fill(B.data,0); for (int i = 0; i <cols; i++) { for (int j = 0; j <=i; j++) { B.data[i*cols+j] += A.data[i]*A.data[j]; } for (int k = 1; k < A.numRows; k++) { int indexRow = k*cols; double valI = A.data[i+indexRow]; int indexB = i*cols; for (int j = 0; j <= i; j++) { B.data[indexB++] += valI*A.data[indexRow++]; } } } }
java
public static void pow(ComplexPolar_F64 a , int N , ComplexPolar_F64 result ) { result.r = Math.pow(a.r,N); result.theta = N*a.theta; }
java
public static void sqrt(Complex_F64 input, Complex_F64 root) { double r = input.getMagnitude(); double a = input.real; root.real = Math.sqrt((r+a)/2.0); root.imaginary = Math.sqrt((r-a)/2.0); if( input.imaginary < 0 ) root.imaginary = -root.imaginary; }
java
public boolean computeDirect( DMatrixRMaj A ) { initPower(A); boolean converged = false; for( int i = 0; i < maxIterations && !converged; i++ ) { // q0.print(); CommonOps_DDRM.mult(A,q0,q1); double s = NormOps_DDRM.normPInf(q1); CommonOps_DDRM.divide(q1,s,q2); converged = checkConverged(A); } return converged; }
java
private boolean checkConverged(DMatrixRMaj A) { double worst = 0; double worst2 = 0; for( int j = 0; j < A.numRows; j++ ) { double val = Math.abs(q2.data[j] - q0.data[j]); if( val > worst ) worst = val; val = Math.abs(q2.data[j] + q0.data[j]); if( val > worst2 ) worst2 = val; } // swap vectors DMatrixRMaj temp = q0; q0 = q2; q2 = temp; if( worst < tol ) return true; else if( worst2 < tol ) return true; else return false; }
java
public void setup( int numSamples , int sampleSize ) { mean = new double[ sampleSize ]; A.reshape(numSamples,sampleSize,false); sampleIndex = 0; numComponents = -1; }
java
public double[] getBasisVector( int which ) { if( which < 0 || which >= numComponents ) throw new IllegalArgumentException("Invalid component"); DMatrixRMaj v = new DMatrixRMaj(1,A.numCols); CommonOps_DDRM.extract(V_t,which,which+1,0,A.numCols,v,0,0); return v.data; }
java
public double[] sampleToEigenSpace( double[] sampleData ) { if( sampleData.length != A.getNumCols() ) throw new IllegalArgumentException("Unexpected sample length"); DMatrixRMaj mean = DMatrixRMaj.wrap(A.getNumCols(),1,this.mean); DMatrixRMaj s = new DMatrixRMaj(A.getNumCols(),1,true,sampleData); DMatrixRMaj r = new DMatrixRMaj(numComponents,1); CommonOps_DDRM.subtract(s, mean, s); CommonOps_DDRM.mult(V_t,s,r); return r.data; }
java
public double[] eigenToSampleSpace( double[] eigenData ) { if( eigenData.length != numComponents ) throw new IllegalArgumentException("Unexpected sample length"); DMatrixRMaj s = new DMatrixRMaj(A.getNumCols(),1); DMatrixRMaj r = DMatrixRMaj.wrap(numComponents,1,eigenData); CommonOps_DDRM.multTransA(V_t,r,s); DMatrixRMaj mean = DMatrixRMaj.wrap(A.getNumCols(),1,this.mean); CommonOps_DDRM.add(s,mean,s); return s.data; }
java
public double response( double[] sample ) { if( sample.length != A.numCols ) throw new IllegalArgumentException("Expected input vector to be in sample space"); DMatrixRMaj dots = new DMatrixRMaj(numComponents,1); DMatrixRMaj s = DMatrixRMaj.wrap(A.numCols,1,sample); CommonOps_DDRM.mult(V_t,s,dots); return NormOps_DDRM.normF(dots); }
java
public static <T extends DMatrix> boolean decomposeSafe(DecompositionInterface<T> decomp, T M ) { if( decomp.inputModified() ) { return decomp.decompose(M.<T>copy()); } else { return decomp.decompose(M); } }
java
public static DMatrix2 extractColumn( DMatrix2x2 a , int column , DMatrix2 out ) { if( out == null) out = new DMatrix2(); switch( column ) { case 0: out.a1 = a.a11; out.a2 = a.a21; break; case 1: out.a1 = a.a12; out.a2 = a.a22; break; default: throw new IllegalArgumentException("Out of bounds column. column = "+column); } return out; }
java
public static boolean decomposeSafe(DecompositionInterface<ZMatrixRMaj> decomposition, ZMatrixRMaj a) { if( decomposition.inputModified() ) { a = a.copy(); } return decomposition.decompose(a); }
java
public static void invert( final int blockLength , final boolean upper , final DSubmatrixD1 T , final DSubmatrixD1 T_inv , final double temp[] ) { if( upper ) throw new IllegalArgumentException("Upper triangular matrices not supported yet"); if( temp.length < blockLength*blockLength ) throw new IllegalArgumentException("Temp must be at least blockLength*blockLength long."); if( T.row0 != T_inv.row0 || T.row1 != T_inv.row1 || T.col0 != T_inv.col0 || T.col1 != T_inv.col1) throw new IllegalArgumentException("T and T_inv must be at the same elements in the matrix"); final int M = T.row1-T.row0; final double dataT[] = T.original.data; final double dataX[] = T_inv.original.data; final int offsetT = T.row0*T.original.numCols+M*T.col0; for( int i = 0; i < M; i += blockLength ) { int heightT = Math.min(T.row1-(i+T.row0),blockLength); int indexII = offsetT + T.original.numCols*(i+T.row0) + heightT*(i+T.col0); for( int j = 0; j < i; j += blockLength ) { int widthX = Math.min(T.col1-(j+T.col0),blockLength); for( int w = 0; w < temp.length; w++ ) { temp[w] = 0; } for( int k = j; k < i; k += blockLength ) { int widthT = Math.min(T.col1-(k+T.col0),blockLength); int indexL = offsetT + T.original.numCols*(i+T.row0) + heightT*(k+T.col0); int indexX = offsetT + T.original.numCols*(k+T.row0) + widthT*(j+T.col0); blockMultMinus(dataT,dataX,temp,indexL,indexX,0,heightT,widthT,widthX); } int indexX = offsetT + T.original.numCols*(i+T.row0) + heightT*(j+T.col0); InnerTriangularSolver_DDRB.solveL(dataT,temp,heightT,widthX,heightT,indexII,0); System.arraycopy(temp,0,dataX,indexX,widthX*heightT); } InnerTriangularSolver_DDRB.invertLower(dataT,dataX,heightT,indexII,indexII); } }
java
public static DMatrixRMaj[] span(int dimen, int numVectors , Random rand ) { if( dimen < numVectors ) throw new IllegalArgumentException("The number of vectors must be less than or equal to the dimension"); DMatrixRMaj u[] = new DMatrixRMaj[numVectors]; u[0] = RandomMatrices_DDRM.rectangle(dimen,1,-1,1,rand); NormOps_DDRM.normalizeF(u[0]); for( int i = 1; i < numVectors; i++ ) { // System.out.println(" i = "+i); DMatrixRMaj a = new DMatrixRMaj(dimen,1); DMatrixRMaj r=null; for( int j = 0; j < i; j++ ) { // System.out.println("j = "+j); if( j == 0 ) r = RandomMatrices_DDRM.rectangle(dimen,1,-1,1,rand); // find a vector that is normal to vector j // u[i] = (1/2)*(r + Q[j]*r) a.set(r); VectorVectorMult_DDRM.householder(-2.0,u[j],r,a); CommonOps_DDRM.add(r,a,a); CommonOps_DDRM.scale(0.5,a); // UtilEjml.print(a); DMatrixRMaj t = a; a = r; r = t; // normalize it so it doesn't get too small double val = NormOps_DDRM.normF(r); if( val == 0 || Double.isNaN(val) || Double.isInfinite(val)) throw new RuntimeException("Failed sanity check"); CommonOps_DDRM.divide(r,val); } u[i] = r; } return u; }
java