The functions (k=1,2) are stored in a 5-dimensional array FF(k,n,i,j,l) (n=1,MPH), (i=0,MY), (j=0,MXI), (l=0,MLM). The integral over y from 0 to is stored in FINT(k,n,i,j,l). The integral over the full range is then FINT(k,n,MY,j,l) For integration, the trapezondal rule is used, which means the function is approximated by a piecewise linear function. The sum of FINT(k,n,MY,j,l) over n is stored in FALL(k,j,l).
For a given initial condition, calculate the parameters and
and find FALL(*) by 2-dimensional interpolation.
(The asterisk * indicates the appropriate sum over the
initial polarization, i,e.,
FALL(*)=FALLFALL).
Then, calculate the total probability P (eq.(131) times the time interval DT):
Generate a uniform random number in the interval (0,1).
If , decide to emit a photon and, otherwise reject.
If rejected, the helicity of the electron should be changed, according to
eq.(12), to
If accepted, decide how many laser photons to absorb. To do so, sum up FINT(*,n,MY,j,l) from n=1 to until the sum becomes larger than . Then, will be the number of photons.
Once is determined, the photon energy is determined by
where is another uniform random number. The left hand side is known
for the mesh point of y (i.e., FINT(k,n,MY,j,l)).
Since we approximate
by a piecewise linear function of y, the left hand side is a
quadratic function between successive y's.
Thus, inverse interpolation with respect to i by solving a
quadratic equation gives the photon energy to be emitted.
The helicities of the final photon and electron are calculated from
for . This is done by directly calling a Bessel function routine.