How can I truncate a random generated exponentail distribution?

Hi, I'm trying to truncate an exponential distribution generated with the random function:

for mu_curr=1:100;
  mu_exp=linspace(0.002,1,100);
    x=1:desired_length;
    mu_curr_exp=mu_exp(curr_distr);
    %% generating ISI
    ISI_exp=((random('Exponential',mu_curr_exp,length(x),1)));
    %% concatenating in a matrix
    Y_EXP(curr_distr,1:length(x))=ISI_exp';
end

I am using this loop because I have to change the mu value 100 times. I have to truncate this so number don't go below 1e-6 and above 0.2. Do you have any ideas? I tried to use the distribution object but I don't know how to use it to perform calculations on the results. Thank you,

Carolina

Answer 1

[JESUS DAVID ARIZA ROYETH]

I leave you an idea that can be useful, in R numbers are generated with an exponential distribution in the range you want :

n=20;%amount of values to generate
mu=0.1;%mu
R = mapminmax(exprnd(mu,1,n),1e-6,0.2)

[Carolina Scandellari]

Thank you, I think I've figured out how to solve it.

Answer 2

[KALYAN ACHARJYA]

I have to truncate this so number don't go below 1e-6 and above 0.2. Do you have any ideas?

Is this?

I am expecting the number is Y_EXP, if else, please replace Y_EXP with respective number.

Y_EXP(Y_EXP<1e-6 | Y_EXP>0.2)=0;

[Carolina Scandellari]

Thank you, I think I've figured out how to solve it.

Answer 3

[Star Strider]

Use the truncate function (introduced in R2013a).

See the documentation section on pd to use it with different distributions.

[Carolina Scandellari]

Yes you can use truncate but you will create an object, and I can't perform any calculations on objects. Thank you anyway

[Star Strider]

My pleasure.

However, I disagree!

pd = makedist('Normal')
t = truncate(pd,0,inf)
y = pdf(t, (0:5:10))

produces:

y =
         0.797884560802865       2.9734390294686e-06      1.53891972534128e-22

What ‘calculations’ do you want to perform with it?

See: Working with Probability Distributions and specifically Probability Distribution Objects for an extended discussion of their capabilities.