Contents

%Monte Carlo simulation sampling notes
%sampling_notes.m    VERSION 1.0      7 January 2014

seed random number generator to state zero (debugging and development)

utility function: seedRNG0.m
if verLessThan('matlab', '8.0.0')
    RandStream.setDefaultStream(RandStream('mt19937ar','seed',0));
else
    RandStream.setGlobalStream(RandStream('mt19937ar','seed',0));
end

seed random number generator using system date/time (simulation)

utility function: seedRNG.m
if verLessThan('matlab', '8.0.0')
    RandStream.setDefaultStream(RandStream('mt19937ar','seed',sum(clock*100)));
else
    RandStream.setGlobalStream(RandStream('mt19937ar','seed',sum(clock*100)));
end

draw N samples from rectangular distribution ranging from 0.0 - 1.0

N=5; sample = rand(N,1)

sample =

    0.6685
    0.8842
    0.4474
    0.9790
    0.5871

generate N values drawn from uniform distribution in the range of [a, b]

N=10; a=1; b=10;
sample = a + (b-a) .* rand(N,1)
%
%round results to the nearest whole number
sample = floor(sample + 0.5)

sample =

    6.0448
    5.3492
    7.1893
    8.8235
    5.6970
    2.5136
    6.8506
    3.2634
    5.3663
    9.0789


sample =

     6
     5
     7
     9
     6
     3
     7
     3
     5
     9

generate random integers from rectangular discrete distribution

ranging from 1 to iMax
iMax=10;  nSamples=5;
sample = randi(iMax, nSamples, 1)

sample =

     7
     7
     8
     2
    10

generate random numbers drawn from the standard normal distribution

nSamples = 5;  sample = randn(nSamples,1)
%
% generate random numbers drawn gaussian dist(mean=mu, std dev=sigma)
mu=0; sigma=1; nSamples=5;
sample = mu + sigma .* randn(nSamples,1)

sample =

    1.3959
    0.9430
   -0.5264
    2.2004
    1.1068


sample =

   -0.3275
   -0.3259
   -0.1168
   -0.5526
   -0.2399

draw a sample from a normal distribution of specified mean; std dev

mu=100; sigma=15; nSamples=15;
sample = normrnd(mu, sigma, nSamples, 1)

sample =

   73.4265
   89.0217
  103.9900
   99.8874
   73.9950
  107.5702
   92.7374
  107.9225
   69.6347
  120.0430
   90.5218
  128.2619
   96.3719
   93.4447
  100.7020

draw a random sample from a finite, user-defined population

(with or without replacement)
population = [0 0 0 0 0 0 0 0 0 1 1]; nSamples=10;
bReplacement = true; %specifiy sampling with replacement
sample = randsample(population, nSamples, bReplacement)
%
% draw samples from a population of amino acids (i.e., 'ACGT') and
% specify the a priori probability of each member of the population
mySequence = randsample('ACGT', 24, true, [0.15 0.35 0.35 0.15])

sample =

     1     0     0     0     0     0     0     0     0     1


mySequence =

GCGTGCAGGCAGGCGCCGTGCGGT

generate a random permutation for the numbers 1 through N

N=10;
sample = randperm(N)

sample =

     6     9    10     4     2     3     7     8     1     5

shuffle the order of the elements in a row vector

data = [10 20 30 40 50 60 70 80 90 100]
order = randperm( length(data) )
shuffled_data = data(order)

data =

    10    20    30    40    50    60    70    80    90   100


order =

     8     5     1     3     7    10     6     2     9     4


shuffled_data =

    80    50    10    30    70   100    60    20    90    40

Roll a die

outcome = randi([1 6],1)

outcome =

     2

Flip a coin

outcome = rand < 0.5
%or
outcome = randi([0 1],1)
%
%flip a coin N times
N=10;
outcome = rand(N,1) < 0.5
%or
outcome = randi([0 1],N,1)

outcome =

     1


outcome =

     1


outcome =

     0
     0
     1
     0
     1
     1
     1
     0
     0
     0


outcome =

     0
     0
     0
     1
     1
     1
     0
     0
     0
     1


Published with MATLAB® 7.7