normrnd

Generating random number with normal distribution

Usages

normrnd(mu, sigma)

• It returns pseudo-random numbers of normal distributions with means and standard deviations given by mu and sigma, respectively.
• mu and sigma should be arrays with the same sizes, or one of them should be scalar.
• If one of the inputs is a scalar, it will be expanded to match the size of the other input.
• sigma should contain non-negative numbers.

normrnd(mu, sigma, m)

• It returns an m-by-m array containing pseudo-random numbers of normal distributions with means and standard deviations given by mu and sigma, respectively.
• Each of mu and sigma should be a scalar or should have size [m, m].
• If mu or sigma is a scalar, it will be expanded to have the size [m, m].
• m should be a non-negative integer.
• sigma should contain non-negative numbers.

normrnd(mu, sigma, s1, s2,..., sn)

• It returns an array of pseudo-random numbers of normal distributions with means and standard deviations given by mu and sigma, respectivley.
• s1, s2, ... sn should be real, non-negative integers.
• The output array has the size [s1, s2, ..., sn].
• Each of mu and sigma should be either a scalar or should have the size [s1, s2, ..., sn].
• If mu or sigma is a scalar, it will be expanded to match the size [s1, s2, ..., sn].
• sigma should contain non-negative numbers.

normrnd(mu, sigma, sizeVec)

• It returns an array of pseudo-random numbers of normal distributions with means and standard deviations given by mu and sigma.
• The output has the size specified by the vector sizeVec.
• Each of mu and sigma should be a scalar or should have size sizeVec.
• If mu or sigma is a scalar, it will be expanded to match the size sizeVec.
• sigma should contain non-negative numbers.

Discussions

• Since this app's algorithms for generating random numbers and initial states may be different from MATLAB's, users should not expect this app and MATLAB to give the same sequences of pseudo-random numbers.

Examples

• In this example, we generate 100,000 random samples by normrnd with $\mu=1$ and $\sigma=2$. Then, we estimate the mean and standard deviation from the samples. We find that they are indeed accurate with $\mu\approx 1$ and $\sigma\approx 2$.

Input

% Vector of 100,000 numbers from normal distribution with mean equal to 1 and SD equal to 2.
r = normrnd(1, 2, 1, 100000);
% Mean estimated from r.
mean(r)
% Standard deviation estimated from r.
sqrt(mean((1-r).^2))

Output

ans = 
 1.0011

ans = 
 2.0023