           How to Use "lslp"
           -----------------
      J.E. Boyd - UCSD - 21-Feb-97

Purpose:
--------

"lslp" uses least-squares linear prediction to determine the
fundamental frequency of a signal and it's phase relative to a cos
function.

Use:
----

Prepare an input file consisting of the signal values in ascii format. 
Include no other numbers in the input file.

Call "lslp" as follows:

	lslp [command switches] <input file name>

"lslp" will print the following information to stderr:

	- Number of samples read from the input file
	- Fundamental frequency and how it is derived (see switches below)
	- Magnitude and Phase of the signal

Command line switches modify the behavior of "lslp" as follows:

	-l <on/off>	use least-squares linear prediction to find the
                        fundamental frequency (default on).
	-f float	if -l off is specified (see above) use the value
                        specified here as the fundamental frequency
                        default (0.06)
	-m int		when -l on, this number specifies the order of
                        the least-squares fit (default 5)

	The following switches force "lslp" to print the specified
        information to stdout (all are default off):

	-spect <on/off>	power spectrum, only works for -l on
	-phase <on/off>	signal phase
	-mag <on/off>	signal magnitude
	-data <on/off>	signal with the background removed
	-fit <on/off>	the sinusoid that best fits the data
	-freq <on/off>	the frequency of the sinusoid

	The above switches allow the user to aquire whatever information
	is needed from the input signal.

Notes:
------

1.  I compiled this using gcc so there may be some C++ style comments using
    the "//" convention that will confound an ANSI C compiler.

2.  The program uses the following routines/files from Numerical Recipes in C:
      complex.c
      complex.h
      laguer.c
      lubksb.c
      ludcmp.c
      array allocation routines
    I do not own the copyright for that code so you must acquire it yourself
    in order to compile lslp.
