Probability of being under a target

Hi,

I have a result variable X that is defined by an equation of random variables.

We used prctile to see the results we get with several probabilities, but now we need to known what is the probability that X is going to be under a certain value.

Is there any function that can give this?

One way I was thinking about doing this is to count the number of samples in the simulation where X is below the mark, but I don't known how to do this neither.

Thank you very much,

Answer 1

[Matt Tearle]

Counting the number of samples with a given criterion is easy enough:

threshhold = pi;
nnz(X < threshhold)

The ecdf (empirical CDF) function might also help.

[Alexandra]

Thank you, that was great.

I tried this:

Prob = mean(bsxfun(@lt,X,threshold),1)

Is this correct? The simulation is already running. Otherwise I will just try to do what you suggested.

[Matt Tearle]

So X is a matrix, where each row is a simulation, and each column is...? Then you have a different threshold for each column? In that case, yes, what you have seems right to me -- it will give the proportion of successful trials (X value less than threshold) for each different threshold value.

[Alexandra]

Only one column, the X column. Each row is a simulation. I think it worked. Thank you very much.

[Matt Tearle]

For a single column bsxfun is overkill. mean(X < threshold) would do it.