Re: [Gnuplot-bugs] new assign() function.

[Ethan A M. <merritt@u.washington.edu>]

On Saturday 02 June 2007 07:17, pl...@pi... wrote:
> 
> The problem occurs in using replot. Since the line defining the initial  
> conditions only gets executed when loading the .gnu file, calling replot  
> here gives me an incorrect result by retaining that last values of started  
> and hence prev_x, prev_y :

Your examples are so complicated that I can't see what is supposed to happen.
But let me point out that it is perfectly legal to put the definition
inside the plot statement, in which case I think it will be re-executed on 
every "replot".

  plot started=0, foo(x), baz(x)

	Ethan


> 
> A work around would be to plot some dummy data before the main plot that  
> called another function where the variables are initialised. In my case it  
> is sufficient to reassign started=0 as shown. There may be a more elegant  
> solution.
> 
> 
> #started=0;
> reset(x)=assign("started",0.0); # called by dummy plot to ensure  
> initialisation on replot
> add_aug(x,y)=assign("area",(started>0)?\
>                                        area + (x-prev_x)*(y+prev_y)/2.0  
> + 0.0*(assign("prev_x",x) + assign("prev_y",y) ) \
>                                      :  0.0*( assign("started",1)  
> + assign("prev_x",x) + assign("prev_y",y) ) \
>                     ) ;
> 
> plot  "init.data"  using 1:(reset(0)) t "" \
> ,  datafile   using  
> 1:(add_aug(strptime("%H:%M",stringcolumn(1))/3600,P( Th4($3-2.0) - Th5($2)  
> ))/10) axes x1y2  with lines s f t "Energy*100/kWh"
> 
> 
> best regards/
> 
> 

-- 
Ethan A Merritt
Biomolecular Structure Center
University of Washington, Seattle 98195-7742