next up previous contents
Next: Function: time_freq_map() Up: GRASP Routines: Time-Frequency Methods Previous: Steger's Line Detection Routines   Contents


Structure: struct struct_tfparam

The structure struct_tfparam is the main structure used by the time-frequency routines and the program calling these routines. Some of the fields of this structure are used by the time-frequency routines while others are others are useful for book keeping. (We will use the abbreviation BK to denote the fields used for Book Keeping.)

The fields of this structure are:
struct_tfparam{

int run_number: BK. An identification label.
float f_lower: BK. The lower frequency cutoff for the signals injected.
int start_segment: BK. Used by the calling program to identfy the first data segment to analyse.
int transformtype: The type of the time-frequency transform to be used to construct the map. Should be set to any one of the three Macros defined in file grasp.h, namely, WIGNERTF, WIGNERTF_NP, WFFTWTF, or CHOIWILLIAMS which correspond to the currently implemented transforms namely,the Wigner-Ville transform with zero padding, Wigner-Ville transform without zero padding, the windowed Fourier transform and the Choi-Williams transform respectively.
int windowidth: The width (in number of data points) of the window to use for the Choi-Williams and the windowed Fourier transform.
int offset_step_size: This variable governs the resolution at which the TF map is computed and should normally be set to unity. If this variable is not set to unity then the TF distributions are not computed for every value of time. For example if this variable is set to 2 then the TF distribution is computed at half the resolution.
int num_of_segments: BK. The number of data segments to analyse.
float maxpixelval: Can be used to set a threshhold on the values of the pixels in the TF map. This value can be computed using the routine compute_scalefactor().
int DIM: The data dimension of the data segment array.
int ND: The data dimension of the subsegment array.
int PD: The dimension of the TF map. Must be less that ND/4.
int PRE: The number of data points to skip at the beginning of each segment.
int POST: The number of data points to skip at the end of each segment.
int TD: Has to be set to ND/PD.
int FD: Has to be set to ND/(4*PD)
float rescale_factor: Used to compute maxpixelval via the routine compute_scale_factor(). Has to be set to a number greater than unity.
float hscale: BK. An arbitrary scaling number.
int noisetype: BK. The type of noise.
float srate: The sampling rate.
}


next up previous contents
Next: Function: time_freq_map() Up: GRASP Routines: Time-Frequency Methods Previous: Steger's Line Detection Routines   Contents
Bruce Allen 2000-11-19