public void copyHtoD() {
if (System.getProperty("use_cuda").equals("true")) {
if (this.persist == false) this.persist = true;
if (this.cPointer != null) {
SimpleCuBlas.free(this.cPointer);
// JCublas.cublasInit();
// JCublas.cublasFree(this.cPointer);
this.cPointer = null;
}
this.cPointer = SimpleCuBlas.alloc(this.data());
}
}
public DMatrix sub(DMatrix other) {
assert this.length() == other.length()
: System.out.printf("Length is not equal. %d - %d\n", this.length(), other.length());
DMatrix m = new CUDAMatrix(this.rows(), this.columns(), this.toArray());
SimpleCuBlas.axpy(-1.0, other, m);
return m;
}
public void copyDtoH() {
if (System.getProperty("use_cuda").equals("true")) {
if (this.cPointer != null) {
// JCublas.cublasInit();
SimpleCuBlas.getData(this, this.cPointer, Pointer.to(this.data()));
}
}
}
public static DMatrix randn(int r, int c, boolean _persist) {
DMatrix m = new CUDAMatrix(r, c);
m.persist = _persist;
for (int i = 0; i < r * c; i++) m.put(i, RandomUtils.nextGaussian());
if (m.persist) {
m.cPointer = SimpleCuBlas.alloc(m.data());
}
return m;
}
public void close() {
// System.err.printf("close() in CUDAMatrix\n");
if (this.cPointer != null) {
// System.out.printf("Releasing the CUDA Pointer\n");
this.persist = false;
SimpleCuBlas.free(this.cPointer);
// JCublas.cublasInit();
// JCublas.cublasFree(this.cPointer);
this.cPointer = null;
}
}
// result = this*other
public DMatrix mmul(boolean tA, boolean tB, DMatrix B, DMatrix C) {
int m = tA ? this.columns() : this.rows();
int n = tB ? B.rows() : B.columns();
int k = tA ? this.rows() : this.columns();
int kB = tB ? B.columns() : B.rows();
assert (k == kB);
if (C.rows != m || C.columns != n) {
if (C != this && C != B) {
C.resize(m, n);
} else {
System.err.printf(
"[ALERT] Should not resize result matrix because it is used in-place. But doing it anyway.\n");
}
}
if (C == this || C == B) {
/* actually, blas cannot do multiplications in-place. Therefore, we will fake by
* * allocating a temporary object on the side and copy the result later.
* */
DMatrix temp = new CUDAMatrix(m, n);
if (m == 1) {
SimpleCuBlas.gemv(tB, B, this, temp, 1.0, 0.0);
} else {
SimpleCuBlas.gemm(tA, tB, this, B, temp, 1.0, 0.0);
}
if (temp.rows() == C.rows() && temp.columns() == C.columns()) SimpleCuBlas.copy(temp, C);
else {
C.resize(m, n);
SimpleCuBlas.copy(temp, C);
}
} else {
if (m == 1) {
SimpleCuBlas.gemv(tB, B, this, C, 1.0, 0.0);
} else {
SimpleCuBlas.gemm(tA, tB, this, B, C, 1.0, 0.0);
}
}
return C;
}
public DMatrix sumRows(int startRow, int howMany) {
DMatrix sum = DMath.createMatrix(1, this.columns());
DMatrix multiplier = DMath.createOnesMatrix(1, howMany);
SimpleCuBlas.cust_gemv(false, this, multiplier, sum, 1.0, 0.0, startRow, howMany);
return sum;
}
public DMatrix sumRows() {
DMatrix sum = DMath.createZerosMatrix(1, this.columns());
DMatrix multiplier = DMath.createOnesMatrix(1, this.rows);
SimpleCuBlas.gemv(false, this, multiplier, sum, 1.0, 0.0);
return sum;
}
public DMatrix fillWithArray(DMatrix other) {
assert (this.length() % other.length() == 0);
SimpleCuBlas.fillWithArray(other, this);
return this;
}
public DMatrix invi() {
SimpleCuBlas.inverseElements(this);
return this;
}
public DMatrix addi(DMatrix other) {
assert (this.length() == other.length());
// System.out.printf("Using cuda blas\n");
SimpleCuBlas.axpy(1.0, other, this);
return this;
}
public DMatrix mulRowsi(DMatrix colVector) {
assert (this.rows() == colVector.rows() && colVector.columns() == 1);
SimpleCuBlas.mulRows(colVector, this);
return this;
}
public DMatrix muli(DMatrix other) {
assert (this.length() == other.length());
SimpleCuBlas.mul(this, other, this);
return this;
}
public DMatrix mul(DMatrix other) {
assert (this.length() == other.length());
DMatrix m = new CUDAMatrix(this.rows(), this.columns());
SimpleCuBlas.mul(this, other, m);
return m;
}
public DMatrix sqrti() {
SimpleCuBlas.sqrt(this);
return this;
}
public DMatrix sqrt() {
DMatrix m = new CUDAMatrix(this.rows, this.columns(), this.toArray());
SimpleCuBlas.sqrt(m);
return m;
}
public DMatrix subi(DMatrix other) {
assert (this.length() == other.length());
SimpleCuBlas.axpy(-1.0, other, this);
return this;
}
// y = a*X+b
public DMatrix addi(double a, DMatrix other) {
SimpleCuBlas.axpy(a, other, this);
return this;
}
public DMatrix sumColumns() {
DMatrix sum = DMath.createMatrix(this.rows, 1);
DMatrix multiplier = DMath.createOnesMatrix(this.columns, 1);
SimpleCuBlas.gemm(false, false, this, multiplier, sum, 1.0, 0.0);
return sum;
}
public DMatrix mulRows(DMatrix colVector) {
assert (this.rows() == colVector.rows() && colVector.columns() == 1);
DMatrix m = new CUDAMatrix(this.rows(), this.columns(), this.toArray());
SimpleCuBlas.mulRows(colVector, m);
return m;
}
public void updateDeviceData() {
if (this.cPointer != null) SimpleCuBlas.updateData(this.cPointer, this.data);
}
// y = 1*x+y
public DMatrix add(DMatrix other) {
assert (this.length() == other.length());
DMatrix m = new CUDAMatrix(this.rows, this.columns, this.data());
SimpleCuBlas.axpy(1.0, other, m);
return m;
}
public DMatrix powi(double v) {
SimpleCuBlas.pow(this, v);
return this;
}
public DMatrix muli(double v) {
SimpleCuBlas.scal(this, v);
return this;
}
public DMatrix inv() {
DMatrix m = DMath.createMatrix(this.rows(), this.columns(), this.toArray());
SimpleCuBlas.inverseElements(m);
return m;
}
public void updateDeviceData(double[] newData) {
if (this.cPointer != null) SimpleCuBlas.updateData(this.cPointer, newData);
}
public DMatrix pow(double v) {
DMatrix m = new CUDAMatrix(this.rows(), this.columns(), this.toArray());
SimpleCuBlas.pow(m, v);
return m;
}