How to add two empirical CDFs without using a for loop?

How can I add two CDFs together if their respective double arrays are of different sizes?

Take for example:

[Fp,Xp] = ecdf(distribution1); % blue CDF shown below
[Fv,Xv] = ecdf(distribution2); % orange CDF shown below

size(Fp,1)
ans = 492
size(Xp,1)
ans = 492

size(Fv,1)
ans = 899
size(Pv,1)
ans = 899

If Fp is the probability of event p happening at time Xp, and Fv is the probability of event v happening at time Xv, how could I calculate Fpv and Xpv if the double arrays are of unequal sizes? I hope this makes sense, sorry if it's a confusing explanation of what I'm trying to do.

The formula I want to use is:

Fpv(x) = Fp(x) + Fv(x) - Fp(x)*Fv(x)

Which assumes that Fp(x)*Fv(x) < Fp(x) + Fv(x)

[Mirlan Karimov]

Why not interpolating Fv in 899 points. interp(Fv, 2)

or other way around: interpolating Fp in 492 points

[Rik]

I don't think you have any option that doesn't involve resampling one of your arrays. I haven't read the doc for interp, but there should be an option to use a spline.

Answer 1

[Darren Kenney]

Solution:

times = 0:0.1:2000;

[Fp,Xp] = ecdf(distribution1);
[uniqueXp, uniqueXpIndex] = unique(Xp); 
interpolatedP = interp1(uniqueXp, Fp(uniqueXpIndex), times, 'previous');

[Fv,Xv] = ecdf(distribution2);
[uniqueXv, uniqueXvIndex] = unique(Xv);
interpolatedV = interp1(uniqueXv, Fv(uniqueXvIndex), times, 'previous');