Next: Function: generate_cubic
Up: GRASP Routines: Template Bank
Previous: Example: match_fit program
Contents
Structure: struct cubic_grid
This structure is used to store precomputed coefficients of the cubic
fit to the match function, generated by match_cubic() on an
equally spaced grid in the parameter space. The stucture
stores the coefficients as well as all the information required to
generate, retrieve, and interpolate among them. The fields of this
structure are:
struct cubic_grid {
 int n;
The number of points along the side of the grid.
 float m_mn;
The minimum mass of an object in the parameter space covered by the
grid (solar masses).
 float m_mx;
The minimum mass of an object in the parameter space covered by the
grid (solar masses).
 float dm;
The spacing between grid points (solar masses); equal to
.
 float match;
The match level (between 0 and 1) out to which the cubic fit was
made.
 float angle;
The angle (radians) counterclockwise from the axis to the
axis (see below).
 int order;
Twice the postNewtonian order of the chirp templates used to
compute the match function.
 float srate;
The sampling rate of the chirp templates used to compute the match
function.
 float flo;
The initial frequency of the chirp templates used to compute the
match function.
 float ftau;
The reference frequency used to define the
coordinates.
 int detector;
The index of the detector site in the data file detectors.dat,
used to identify a noise curve for computing the match function.
 float ***coef;
A pointer to the array of coefficients.
};
The cubic_grid.coef field points to an array of the form coef[0..n1][0..n1][0..9]. The first two indecies [i][j]
identify a point in the mass parameter space:
and
. The third index identifies a
particular coefficient computed at that point. The individual
coefficients are defined as follows: The first 7 entries [0..6]
are the actual coefficients of the cubic fit to the match function
:

(9.12.215) 
where are small displacements in directions at an angle cubic_grid.angle counterclockwise from the
directions, respectively. If one considers only the quadratic part of
this fit, the equation
defines an
ellipse in the plane. Entries [7] and [8] are then
the semimajor and semiminor axes of this ellipse, respectively (in
units of seconds), and entry [9] is the angle (in radians)
counterclockwise from the axis to the semimajor axis. The entries
[7..9] can be computed without too much difficulty from the
entries [0..2] and the value of match, but it can be
useful to have them precomputed.
Next: Function: generate_cubic
Up: GRASP Routines: Template Bank
Previous: Example: match_fit program
Contents
Bruce Allen
20001119