Introduction to Matlab Statistics Toolbox

advertisement
Math 4510/5510
Spring 2014
A tutorial to the Matlab Statistics Toolbox
Math 4510/5510
Lab notes 1.
1. Introduction
The user’s guide can be downloaded here,
http://www.mathworks.com/help/releases/R13sp2/pdf_doc/stats/stats.pdf
The documentation of the Matlab statistics toolbox can be found here,
http://www.mathworks.com/help/stats/index.html;jsessionid=0bf4f3df6d60e53cb7a4efa0eba6
Figure 1. The Statistics Toolbox documentation
We will introduce the most basic functions:
Exploratory data Analsyis
Probability Distributions
Math 4510/5510
Spring 2014
Hypothesis Tests
Regression and ANOVA (optional)
Introduction to Matlab Statistics Toolbox
1. 1. Functions list
(Button on the left in Fig 1. )


Descriptive Statistics
Statistical Visualization
1.2. Classes
(Button at the middle lower corner in Fig 1. )


Probability distractions
Regression and ANOVA – linear regression
2. Examples
Example 1. (descriptive statistics)
Data set simulated stock returns stockreturns.mat
> load stockreturns %load the data
> who
Your variables are:
stocks
> size(stocks) % find the size of the data
> stocks(:, 2) % call the second column of the matrix
% Descriptive statistics , find statistics for the given data
> mean(stocks(:,2); %find the mean of the column 2
> mean_stocks= zeros(10,1); % define a 10X1 matrix, clear memory
> i=1:10; mean_stocks(i, 1)=mean(:,i); % calculate mean for ith column
> mean_stocks
HW Q1. Find standard deviation, minimum, maximum, range, 25, 50, and 75 percentile
Example 2 (visualize data)
> hist(stocks(:,1);
> %Multiple plots on the same panel
> subplot(2,2,1)
> plot
Math 4510/5510
Spring 2014
Lab Notes 2
This lab will introduce
 Data input and out put in Matlab
 Basics of descriptive statistics
Please read Chapter 2 in Text book #2
% recall what last lab
% data set stockreturns.mat
% New data set: gas.mat
% for review, plot hist for gas.mat
% data input and output
% csvwrite(‘yourfile.csv’, variable) and a=csvread(‘yourfile.csv’)
% Exercise 1. Load/Read/write for data set: gas.mat
% matrix format in matlab
> stocks(:,2), % call column 2
> stocks(1,:), % call first row
% Distributions in toolbox
> randtool %generate random variables
Example Normal distribution. Generate a random sample of size 1000.
% Descriptive statistics
% 5 number summary, min, 25 percentile, median, 75 percentile, max, range
% function prctile
> prctile(stock(:,1), 25)
….
> % construct 5 number summary
> …
Math 4510/5510
Spring 2014
Exercise
1. Generate a random sample from exponential distribution, with parameter lamda =2;
2. Find 5 number summary for data in 1;
3. generate histogram, boxplot, and normplot for data in 1:
4. write your data in 1 to a file, expSample.csv
5. read expSample.csv to variable a
Download