E.1 Discrete Distrbutions

Bernoulli \(Ber(p)\)
Parameters: \(0 < p < 1\), probability of success
PMF: \(P[X=1] = p, \; P[X=0] = 1-p\)
Inverse CDF: \(F^{-1}(u) = \text{if}\ (u < p), 1 \ \text{else} \ 0\)
Expected Value: \(E[X] = p\)
Variance: \(Var[X] = p(1-p)\)
Generation: BernoulliRV(probOfSuccess:Double, stream:RNStreamIf)
Spreadsheet Generation: = IF(RAND()\(<\) p, 1, 0)
Modeling: the number of successes in one trial
Binomial \(Binom(n,p)\)
Parameters: \(0 < p < 1\), probability of success, \(n\), number of trials
PMF: \(P[X=x] = \binom{n}{x}p^{x}(1-p)^{n-x} \quad x=0,1,\ldots,n\)
Inverse CDF: no closed form available
Expected Value: \(E[X] = np\)
Variance: \(Var[X] = np(1-p)\)
Generation: BinomialRV(pSuccess:Double, numTrials:Int, stream:RNStreamIf)
Spreadsheet Generation: = BINOM.INV(n,p,RAND())
Modeling: the number of successes in \(n\) trials
Shifted Geometric Shifted Geo(\(p\))
Parameters: \(0 < p < 1\), probability of success
PMF: \(P[X=x] = p(1-p)^{x-1} \quad x=1,2,\ldots,\)
Inverse CDF \(F^{-1}(u) = 1 + \left\lfloor \frac{ln(1 - u)}{ln(1 - p)} \right\rfloor\)
Expected Value: \(E[X] = 1/p\)
Variance: \(Var[X] = (1-p)/p^2\)
Generation: ShiftedGeometricRV(probOfSuccess:Double, stream:RNStreamIf)
Spreadsheet Generation: = \(\text{1 + INT(LN(1-RAND())/LN(1-p))}\)
Modeling: the number of trials until the first success
Negative Binomial Defn. 1 NB1(\(r,p\))
Parameters: \(0 < p < 1\), probability of success, \(r^{th}\) success
PMF: \(P[X=x] = \binom{x-1}{r-1}p^{r}(1-p)^{x-r} \quad x=r,r+1\ldots,\)
Inverse CDF: no closed form available
Expected Value: \(E[X] = r/p\)
Variance: \(Var[X] = r(1-p)/p^2\)
Generation: NegativeBinomialRV(probOfSuccess:Double, numSuccess: Int, stream:RNStreamIf)
Spreadsheet Generation: use convolution of shifted geometric
Modeling: the number of trials until the \(r^{th}\) success
Negative Binomial Defn. 2 NB2(\(r,p\))
Parameters: \(0 < p < 1\), probability of success, \(r^{th}\) success
PMF: \(P[Y=y] = \binom{y+r-1}{r-1}p^{r}(1-p)^{y} \quad y=0,1,\ldots\)
Inverse CDF: no closed form available
Expected Value: \(E[Y] = r(1-p)/p\)
Variance: \(Var[Y] = r(1-p)/p^2\)
Generation: value of NegativeBinomialRV minus 1
Spreadsheet Generation: use convolution of geometric
Modeling: the number of failures prior to the \(r^{th}\) success
Poisson Pois(\(\lambda\))
Parameters: \(\lambda > 0\)
PMF: \(P[X=x] = \frac{e^{-\lambda}\lambda^{x}}{x!} \quad x = 0, 1, \ldots\)
Inverse CDF: no closed form available
Expected Value: \(E[X] = \lambda\)
Variance: \(Var[X] = \lambda\)
Generation: PoissonRV(mean: Double, stream:RNStreamIfc)
Spreadsheet Generation: not available, approximate with lookup table approach
Modeling: the number of occurrences during a period of time
Discrete Uniform DU(\(a, b\))
Parameters: \(a \leq b\)
PMF: \(P[X=x] = \frac{1}{b-a+1} \quad x = a, a+1, \ldots, b\)
Inverse CDF: \(F^{-1}(u) = a + \lfloor(b-a+1)u\rfloor\)
Expected Value: \(E[X] = (b+a)/2\)
Variance: \(Var[X] = \left( \left( b-a+1\right)^2 -1 \right)/12\)
Generation: DUniformRV(min: Int, max:Int, stream:RNStreamIfc)
Spreadsheet Generation: =RANDBETWEEN(a,b)
Modeling: equal occurrence over a range of integers