@(@\newcommand{\W}[1]{ \; #1 \; }
\newcommand{\R}[1]{ {\rm #1} }
\newcommand{\B}[1]{ {\bf #1} }
\newcommand{\D}[2]{ \frac{\partial #1}{\partial #2} }
\newcommand{\DD}[3]{ \frac{\partial^2 #1}{\partial #2 \partial #3} }
\newcommand{\Dpow}[2]{ \frac{\partial^{#1}}{\partial {#2}^{#1}} }
\newcommand{\dpow}[2]{ \frac{ {\rm d}^{#1}}{{\rm d}\, {#2}^{#1}} }@)@This is cppad-20221105 documentation. Here is a link to its
current documentation
.
Gradient of Determinant Using Lu Factorization: Example and Test
// Complex examples should supppress conversion warnings# include <cppad/wno_conversion.hpp>
# include <cppad/cppad.hpp>
# include <cppad/speed/det_by_lu.hpp>
// The AD complex case is used by this example so must// define a specializatgion of LeqZero,AbsGeq for the AD<Complex> casenamespace CppAD {
CPPAD_BOOL_BINARY( std::complex<double> , AbsGeq )
CPPAD_BOOL_UNARY( std::complex<double> , LeqZero )
}
bool JacLuDet(void)
{ bool ok = true;
usingnamespace CppAD;
double eps99 = 99.0 * std::numeric_limits<double>::epsilon();
typedef std::complex<double> Complex;
typedef AD<Complex> ADComplex;
size_t n = 2;
// object for computing determinants
det_by_lu<ADComplex> Det(n);
// independent and dependent variable vectorsCPPAD_TESTVECTOR(ADComplex) X(n * n);
CPPAD_TESTVECTOR(ADComplex) D(1);
// value of the independent variable
size_t i;
for(i = 0; i < n * n; i++)
X[i] = Complex( double(i), -double(i) );
// set the independent variablesIndependent(X);
// compute the determinant
D[0] = Det(X);
// create the function object
ADFun<Complex> f(X, D);
// argument valueCPPAD_TESTVECTOR(Complex) x( n * n );
for(i = 0; i < n * n; i++)
x[i] = Complex( double(2 * i) , double(i) );
// first derivative of the determinantCPPAD_TESTVECTOR(Complex) J( n * n );
J = f.Jacobian(x);
/* f(x) = x[0] * x[3] - x[1] * x[2] */
Complex Jtrue[] = { x[3], -x[2], -x[1], x[0] };
for( i = 0; i < n*n; i++)
ok &= NearEqual( Jtrue[i], J[i], eps99 , eps99 );
return ok;
}