Statistics Routines - Prof. Richard B. Goldstein

# Purpose/routine Input Output Example
[1] Descriptive Statistics a list of numbers are entered
that are separated by comma
sorted, mean, quantiles,
skewness, outliers, etc.
box and whisker diagram
-2,1,6,10,7,4,5,1
yields a mean of 6.125
and 18 is an outlier
[2] Discrete Simulation expected frequencies, no. of
simulations
observed frequencies, bar graph
a:3, b:2, c:5
n=1000
[3] Random Number Generation
select from 1 of 10 different
discrete or continuous distributions
random numbers from the
appropriate distribution
normally distributed
with mean=80, st.dev.=10
[4] Statistical Testing of Randomness
choose from 1 of 5 random number
generation routines
results numerically and visually from
various test of randomness
L'Ecuyer's method
[5] Convolution:
sum of n identical variables
a distribution function and interval
area must be positive but not necessarily
integrate to 1 since scale is used
graph of 2 to 50 summations or
all cases from 1 to 20
x*x on [-1,1]
[6] Convolution of any 2 variables two distribution functions
graph of each of the 2 distributions
and the convoluted distribution
Math.abs(x) on [-4, 4]
Math.exp(-x*x/2) on [-3, 3]
[7] Distributions select from 9 continuous and 3 discrete
distributions - parameters and either
p (probability) or x (quantiles)
sliders* are also used
graph showing p and quantile on the
probability density function
Chi Square distribution
with 36 df, p = 0.95,
yields x = 50.9984..

* sliders work best in Chrome and Opera, fair in Internet Explorer, and slow in Firefox
Return to Statistics Index