Submit Assignment For Help

[email protected]

15.075 Statistical Thinking and Data Analysis

Computer Exercises 1

Due October 13, 2011

 
Instructions: Please solve the following exercises using Matlab. One simple way to present your 
solutions is to copy all of your code and results (plots, numerical answers, etc.) into a Word document, 
which you should submit in class. There is no online submission. 
 
Exercise 1 
Write a Matlab script that calculates the mean and median of a sample of 100 uniform random 
numbers between 0 and 2 and the percentage of points in the sample that are greater than 1. 
 
Solution 
 
Here is the script and the results for my iterations: 
 
clear all;
clc;

n = 100;

% Generate 100 uniform random numbers between 0 and 2

x = 2*rand(n,1);

% Calculate the mean

mu = mean(x);

% Calculate the median

med = median(x);

% Find the percentage of points greater than 1

per = sum(x>1)/n;  
 
mu =

1.0089

med =

0.9841

per =

0.4900

 
 
 
 

https://www.statisticshomeworkhelper.com
Exercise 2 
a. Generate a vector of 1000 normal random numbers with mean 8 and variance 25. 
b. Calculate how many elements in the vector are greater than or equal to 9. 
c. What is the sample mean and standard deviation for this sample of 1000 numbers? 
d. What are the 25th and 75th percentiles of the normal distribution with mean 8 and variance 25? 
e. What are the 25th and 75th percentiles of the sample of the 1000 normal random numbers 
generated in part (a)? 
f. Find Φ(0.789) and Φ(‐0.543). (Remember Φ is the cumulative density function for the standard 
normal distribution.) 
 
Solution 
 
Here is the script and the results for my iterations: 
clear all;
clc;

n = 1000;
mu = 8;
sigma = 5;

% a. Generate a vector of 1000 normal random numbers with mean 8 and

variance 25.
x = sigma*randn(n,1)+mu;

% b. Calculate how many elements in the vector are greater than or equal
to 9.
num = sum(x>=9)

% c. What is the sample mean and standard deviation for this sample of
1000 numbers?
sample_mean = mean(x)
sample_std = std(x)

% d. What are the 25th and 75th percentiles of the normal distribution
with mean 8 and variance 25?
q1 = norminv(.25,8,5) % or q1 = 5*norminv(0.25)+8
q3 = norminv(.75,8,5) % or q1 = 5*norminv(0.75)+8

% e. What are the 25th and 75th percentiles of the sample of the 1000
normal random numbers generated in part (a)?
s_q1 = quantile(x,0.25)
s_q3 = quantile(x,0.75)

% f. Find ?(0.789) and ?(-0.543).

normcdf(0.789)
normcdf(-0.543)  
 
num =

428

sample_mean =
 8.0511

sample_std =

4.9974

q1 =

4.6276

q3 =

11.3724

s_q1 =

4.7150

s_q3 =

11.2380

ans =

0.7849

ans =

0.2936
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Exercise 3 
a. Generate a vector of 1000 Poisson random numbers with λ = 2. 
b. Make a histogram and a boxplot of the 1000 numbers from part (a).  
 
  Solution 
 
Here is the script and the plots for my iterations:

clear all;
clc;
close all;

n = 1000;
lambda = 2;

% a. Generate a vector of 1000 Poisson random numbers with ? = 2.

x = poissrnd(lambda,n,1);

% b. Make a histogram and a boxplot of the 1000 numbers from part (a).
figure;
hist(x);

figure;
boxplot(x);

 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Exercise 4 
Answer questions (a) – (c) from 4.36 in your textbook. 
Solution 
 
Here is the script and a scatterplot:

clear all;

close all;

clc;

snowfall = [45 59 82 80 71 60 55 69 79 95];

unemployment = [4.9 5.6 8.5 7.7 7.1 6.1 5.8 7.1 7.6 9.7];

% Scatterplot

plot(snowfall, unemployment, 'x');

grid;

xlabel('Snowfall (inches)');

ylabel('Unemployment (%)');

% Calculates the correlation coefficient

rho = corr(snowfall',unemployment');

 
 

rho = 
 
     0.9835 
 
A strong linear positive relationship is indicated by both the scatter plot and the correlation 
coefficient. This relationship does not mean that the snowfall influences the national unemployment 
rates or vice versa. It means that there is an association between them. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4.44
a.
Below is a time-series plot of the data. There is an increasing trend, so no, the time-series does
not appear to be stationary.

CPI for 1997−1998

164
163
162
CPI

161
160
159

5 10 15 20

Month/Year

b,c. The moving averages and exponentially weighted moving averages are listed below. The MAPE
is 0.2655% and 0.5679% for MA and EWMA respectively.

MA EWMA
159.1000 159.1000
159.3500 159.2000
159.5667 159.3600
159.9333 159.5280
160.1000 159.6424
160.2000 159.7739
160.3000 159.9191
160.5333 160.0953
160.8333 160.3162

2
 161.2000 160.5730

161.4333 160.7584

161.4667 160.8667

161.4667 161.0134

161.6000 161.1907

161.9000 161.3926

162.2000 161.6140

162.5000 161.8512

162.7667 162.0810

163.0000 162.3048

163.2000 162.5238

163.4000 162.7391

163.6667 162.9913

163.8667 163.1930

163.9667 163.3344

d. We have r1 = 0.8694, r2 = 0.7380, and r3 = 0.6149. So CPI in successive months have a rather
strong positive correlation, but the correlation grows weaker with CPI’s removed farther apart.