farith.par.main

su 的源代码库

 FARITH - File ARITHmetic -- perform simple arithmetic with binary files

 farith <infile >outfile [optional parameters]				

 Optional Parameters:							
 in=stdin	input file						
 out=stdout	output file						
 in2=	   second input file (required for binary operations)		
		   if it can't be opened as a file, it might be a scalar
 n=size_of_in,  fastest dimension (used only for op=cartprod is set)	
 isig=		 index at which signum function acts (used only for 	
			op=signum)					
 scale=	value to scale in by, used only for op=scale)		
 bias=		value to bias in by, used only for op=bias)		

 op=noop   noop for out = in						
	   neg  for out = -in						
	   abs  for out = abs(in)					
	   scale for out = in *scale					
	   bias for out = in + bias 					
	   exp  for out = exp(in)					
	   log  for out = log(in)					
	   sqrt for out = (signed) sqrt(in)				
	   sqr  for out = in*in						
	   pinv  for out = (punctuated) 1 / in   			
	   pinvsqr  for out = (punctuated) 1 /in*in 			
	   pinvsqrt for out = (punctuated signed) 1 /sqrt(in) 		
	   add  for out = in + in2					
	   sub  for out = in - in2					
	   mul  for out = in * in2					
	   div  for out = in / in2					
		cartprod for out = in x in2					
		requires: n=size_of_in, fastest dimension in output	
		signum for out[i] = in[i] for i< isig  and			
				= -in[i] for i>= isig			
		requires: isig=point where signum function acts		
 Seismic operations:							
	   slowp   for  out =  1/in - 1/in2	Slowness perturbation	
	   slothp  for  out =  1/in^2 - 1/in2^2   Sloth perturbation	

 Notes:								
 op=sqrt takes sqrt(x) for x>=0 and -sqrt(ABS(x)) for x<0 (signed sqrt)

 op=pinv takes y=1/x for x!=0,  if x=0 then y=0. (punctuated inverse)	

 The seismic operations assume that in and in2 are wavespeed profiles.	
 "Slowness" is 1/wavespeed and "sloth" is  1/wavespeed^2.		
 Use "suop" and "suop2" to perform unary and binary operations on	
 data in the SU (SEGY trace) format.					

 The options "pinvsq" and "pinvsqrt" are also useful for seismic	
 computations involving converting velocity to sloth and vice versa.	

 The option "cartprod" (cartesian product) requires also that the	
 parameter n=size_of_in be set. This will be the fastest dimension	
 of the rectangular array that is output.				

 The option "signum" causes a flip in sign for all values with index	
 greater than "isig"	(really -1*signum(index)).			

 For file operations on SU format files, please use:  suop, suop2	



   AUTHOR:  Dave Hale, Colorado School of Mines, 07/07/89
	Zhaobo Meng added scale and cartprod, 10/01/96
	Zhaobo Meng added signum, 9 May 1997
	Tony Kocurko added scalar operations, August 1997
      John Stockwell added bias option 4 August 2004