Purpose
This is a realistic example using
two levels of AD; see mul_level
.
The first level uses Adolc's adouble type
to tape the solution of an ordinary differential equation.
This solution is then differentiated with respect to a parameter vector.
The second level uses CppAD's type AD<adouble>
to take derivatives during the solution of the differential equation.
These derivatives are used in the application
of Taylor's method to the solution of the ODE.
ODE
For this example the function
@(@
y : \B{R} \times \B{R}^n \rightarrow \B{R}^n
@)@ is defined by
@(@
y(0, x) = 0
@)@ and
@(@
\partial_t y(t, x) = g(y, x)
@)@ where
@(@
g : \B{R}^n \times \B{R}^n \rightarrow \B{R}^n
@)@ is defined by
@[@
g(y, x) =
\left( \begin{array}{c}
x_0 \\
x_1 y_0 \\
\vdots \\
x_{n-1} y_{n-2}
\end{array} \right)
@]@
ODE Solution
The solution for this example can be calculated by
starting with the first row and then using the solution
for the first row to solve the second and so on.
Doing this we obtain
@[@
y(t, x ) =
\left( \begin{array}{c}
x_0 t \\
x_1 x_0 t^2 / 2 \\
\vdots \\
x_{n-1} x_{n-2} \ldots x_0 t^n / n !
\end{array} \right)
@]@
Derivative of ODE Solution
Differentiating the solution above,
with respect to the parameter vector @(@
x
@)@,
we notice that
@[@
\partial_x y(t, x ) =
\left( \begin{array}{cccc}
y_0 (t,x) / x_0 & 0 & \cdots & 0 \\
y_1 (t,x) / x_0 & y_1 (t,x) / x_1 & 0 & \vdots \\
\vdots & \vdots & \ddots & 0 \\
y_{n-1} (t,x) / x_0 & y_{n-1} (t,x) / x_1 & \cdots & y_{n-1} (t,x) / x_{n-1}
\end{array} \right)
@]@
Taylor's Method Using AD
We define the function @(@
z(t, x)
@)@ by the equation
@[@
z ( t , x ) = g[ y ( t , x ) ] = h [ x , y( t , x ) ]
@]@
see taylor_ode
for the method used to compute the
Taylor coefficients w.r.t @(@
t
@)@ of @(@
y(t, x)
@)@.
Memory Management
Adolc uses raw memory arrays that depend on the number of
dependent and independent variables.
The thread_alloc
memory management utilities
create_array
and
delete_array
are used to manage this memory allocation.
// suppress conversion warnings before other includes# include <cppad/wno_conversion.hpp>
//# include <adolc/adouble.h>
# include <adolc/taping.h>
# include <adolc/drivers/drivers.h>
// definitions not in Adolc distribution and required to use CppAD::AD<adouble># include <cppad/example/base_adolc.hpp>
# include <cppad/cppad.hpp>
// ==========================================================================namespace { // BEGIN empty namespace// define types for each leveltypedef adouble a1type;
typedef CppAD::AD<a1type> a2type;
// -------------------------------------------------------------------------// class definition for C++ function object that defines ODEclass Ode {
private:
// copy of a that is set by constructor and used by g(y)CPPAD_TESTVECTOR(a1type) a1x_;
public:
// constructorOde(constCPPAD_TESTVECTOR(a1type)& a1x) : a1x_(a1x)
{ }
// the function g(y) is evaluated with two levels of tapingCPPAD_TESTVECTOR(a2type) operator()
( constCPPAD_TESTVECTOR(a2type)& a2y) const
{ size_t n = a2y.size();
CPPAD_TESTVECTOR(a2type) a2g(n);
size_t i;
a2g[0] = a1x_[0];
for(i = 1; i < n; i++)
a2g[i] = a1x_[i] * a2y[i-1];
return a2g;
}
};
// -------------------------------------------------------------------------// Routine that uses Taylor's method to solve ordinary differential equaitons// and allows for algorithmic differentiation of the solution.CPPAD_TESTVECTOR(a1type) taylor_ode_adolc(
Ode G , // function that defines the ODE
size_t order , // order of Taylor's method used
size_t nstep , // number of steps to takeconst a1type &a1dt , // Delta t for each stepconstCPPAD_TESTVECTOR(a1type) &a1y_ini) // y(t) at the initial time
{
// some temporary indices
size_t i, k, ell;
// number of variables in the ODE
size_t n = a1y_ini.size();
// copies of x and g(y) with two levels of tapingCPPAD_TESTVECTOR(a2type) a2y(n), Z(n);
// y, y^{(k)} , z^{(k)}, and y^{(k+1)}CPPAD_TESTVECTOR(a1type) a1y(n), a1y_k(n), a1z_k(n), a1y_kp(n);
// initialize xfor(i = 0; i < n; i++)
a1y[i] = a1y_ini[i];
// loop with respect to each step of Taylors methodfor(ell = 0; ell < nstep; ell++)
{ // prepare to compute derivatives using a1typefor(i = 0; i < n; i++)
a2y[i] = a1y[i];
CppAD::Independent(a2y);
// evaluate ODE using a2type
Z = G(a2y);
// define differentiable version of g: X -> Y// that computes its derivatives using a1type
CppAD::ADFun<a1type> a1g(a2y, Z);
// Use Taylor's method to take a step
a1y_k = a1y; // initialize y^{(k)}
a1type dt_kp = a1dt; // initialize dt^(k+1)for(k = 0; k <= order; k++)
{ // evaluate k-th order Taylor coefficient of y
a1z_k = a1g.Forward(k, a1y_k);
for(i = 0; i < n; i++)
{ // convert to (k+1)-Taylor coefficient for x
a1y_kp[i] = a1z_k[i] / a1type(k + 1);
// add term for to this Taylor coefficient// to solution for y(t, x)
a1y[i] += a1y_kp[i] * dt_kp;
}
// next power of t
dt_kp *= a1dt;
// next Taylor coefficient
a1y_k = a1y_kp;
}
}
return a1y;
}
} // END empty namespace// ==========================================================================// Routine that tests algorithmic differentiation of solutions computed// by the routine taylor_ode.
bool mul_level_adolc_ode(void)
{ bool ok = true;
double eps = 100. * std::numeric_limits<double>::epsilon();
// number of components in differential equation
size_t n = 4;
// some temporary indices
size_t i, j;
// set up for thread_alloc memory allocatorusing CppAD::thread_alloc; // the allocator
size_t capacity; // capacity of an allocation// the vector x with length n (or greater) in double
double* x = thread_alloc::create_array<double>(n, capacity);
// the vector x with length n in a1typeCPPAD_TESTVECTOR(a1type) a1x(n);
for(i = 0; i < n; i++)
a1x[i] = x[i] = double(i + 1);
// declare the parameters as the independent variable
short tag = 0; // Adolc setup
int keep = 1;
trace_on(tag, keep);
for(i = 0; i < n; i++)
a1x[i] <<= double(i + 1); // a1x is independent for adouble type// arguments to taylor_ode_adolc
Ode G(a1x); // function that defines the ODE
size_t order = n; // order of Taylor's method used
size_t nstep = 2; // number of steps to take
a1type a1dt = 1.; // Delta t for each step// value of y(t, x) at the initial timeCPPAD_TESTVECTOR(a1type) a1y_ini(n);
for(i = 0; i < n; i++)
a1y_ini[i] = 0.;
// integrate the differential equationCPPAD_TESTVECTOR(a1type) a1y_final(n);
a1y_final = taylor_ode_adolc(G, order, nstep, a1dt, a1y_ini);
// declare the differentiable fucntion f : x -> y_final// (corresponding to the tape of adouble operations)
double* y_final = thread_alloc::create_array<double>(n, capacity);
for(i = 0; i < n; i++)
a1y_final[i] >>= y_final[i];
trace_off();
// check function values
double check = 1.;
double t = nstep * a1dt.value();
for(i = 0; i < n; i++)
{ check *= x[i] * t / double(i + 1);
ok &= CppAD::NearEqual(y_final[i], check, eps, eps);
}
// memory where Jacobian will be returned
double* jac_ = thread_alloc::create_array<double>(n * n, capacity);
double** jac = thread_alloc::create_array<double*>(n, capacity);
for(i = 0; i < n; i++)
jac[i] = jac_ + i * n;
// evaluate Jacobian of h at a
size_t m = n; // # dependent variablesjacobian(tag, int(m), int(n), x, jac);
// check Jacobianfor(i = 0; i < n; i++)
{ for(j = 0; j < n; j++)
{ if( i < j )
check = 0.;
else
check = y_final[i] / x[j];
ok &= CppAD::NearEqual(jac[i][j], check, eps, eps);
}
}
// make memroy avaiable for other use by this thread
thread_alloc::delete_array(x);
thread_alloc::delete_array(y_final);
thread_alloc::delete_array(jac_);
thread_alloc::delete_array(jac);
return ok;
}