gnuplot-beta

stat.inc

[Daniel J S. <dan...@ie...>]

I wanted to make use of some functions in 'stat.inc'.  Looking through the file I see there are a few distributions in which p.d.f. and c.d.f. technically should be defined to have a value of zero for negative x.  For example, exponential r.v.s, Weibull r.v.s and Rayleigh r.v.s have a support which is only the non-negative portion of the x-axis.

The way to address this would be to define

#
#define unit step function
#
u_step(x) = x<0 ? 0 : 1

then multiply all the pertinent density/distribution functions by the unit step.

Should I put together a patch for this?

Dan

Re: stat.inc

[<br...@ph...>]

Daniel J Sebald wrote:
> The way to address this would be to define
> 
> #
> #define unit step function
> #
> u_step(x) = x<0 ? 0 : 1
> 
> then multiply all the pertinent density/distribution functions by the 
> unit step.

I think it would be easier to just prefix the existing definitions with

	x<0 ? 0 :

> Should I put together a patch for this?

Sure.