Computes the complimentary error function for real arguments. The erfcx
function takes only a single argument
y = erfcx(x)
where x is either a float or double array. The output
vector y is the same size (and type) as x.
The erfcx function is defined by the integral:
and is an exponentially weighted integral of the normal distribution.
Here is a plot of the erfcx function over the range [-5,5].
--> x = linspace(0,5);
--> y = erfcx(x);
--> plot(x,y); xlabel('x'); ylabel('erfcx(x)');
which results in the following plot.