# CloudRunner

... share your algorithms as a web service

## Algorithm: ISIGNRANK

 Description: ISIGNRANK Bayesian Wilcoxon signed rank sum test based on the Imprecise Dirichlet Process Prob = ISIGNRANK(Y,X) computes the lower and upper posterior probabilityof the hypothesis P(Z>=-Z)>1/2, where Z=Y-X.Prob(1,1) is the lower and Prob(2,1) is the upper.[Prob,H ... (show more) ISIGNRANK Bayesian Wilcoxon signed rank sum test based on the Imprecise Dirichlet Process Prob = ISIGNRANK(Y,X) computes the lower and upper posterior probabilityof the hypothesis P(Z>=-Z)>1/2, where Z=Y-X.Prob(1,1) is the lower and Prob(2,1) is the upper.[Prob,H] = ISIGNRANK(...) performs a hypothesis test for P(Z>=-Z)>1/2with posterior probability 0.95.H=1 indicates that the hypothesis P(Z>=-Z)>1/2 is true with posterior probability greater than 0.95, i.e., Prob(1,1)>0.95. H=0 indicates P(Z>=-Z)>1/2 is not true with posterior greater than 0.95, i.e., Prob(2,1)<0.95. Finally, H=2 indicates an indeterminate instance, i.e., we cannot decide if P(Z>=-Z)>1/2 is true with posterior probability greater than 0.95. This means that the posterior inferences are prior dependent, i.e., Prob(1,1)<0.95 and Prob(2,1)>0.95. [Prob,H] = ISIGNRANK(...,'alpha',ALPHA) returns the result of the hypothesis test with posterior probability gretaer than 1-ALPHA. [Prob,H] = ISIGNRANK(...,'s',sval) sets the value of the prior strength of the Dirichlet Process s to sval. The default value is s=(sqrt(17)-3)/2. For sval=0 it coincides with the Bayesian bootstrap. [Prob,H] = ISIGNRANK(...,'method',M) computes the posterior probability if M is 'exact', or uses a normal approximation if M is 'approximate'. If you omit this argument, ISIGNRANK uses the exact method for small samples and the approximate method for larger samples. [Prob,H] = ISIGNRANK(...,'nsamples',N) sets the number of samples generated from the Dirichlet distribution to compute the posterior probabilities. Default value is 200000. [Prob,H] = ISIGNRANK(...,'display','off') does not show the plot. [Prob,H] = ISIGNRANK(...,'rope',val) introduces a (symmetric) Region of Practical Equivalence (ROPE) around 1/2, i.e., [1/2-val,1/2+val]. [Prob,H] = ISIGNRANK(...,'tail',TAIL) performs the test specified by TAIL: 'right' -- evaluates the hypothesis P(Z>=-Z)>1/2. This is the Bayesian improved version of ranksum(y,x,'tail','right') (default value). 'left' -- evaluates the hypothesis P(Y <= X)>1/2. This is the Bayesian improved version of ranksum(y,x,'tail','left'). 'both' -- performs a two-sided Bayesian test, i.e., H=1 if 1/2 is not included in the 1-ALPHA lower and upper HPD credible intervals. H=0 if 1/2 is included in the 1-ALPHA lower and upper HPD credible. H=2 otherwise, indeterminate case. This is the Bayesian improved version of signrank(y,x). [Prob] = ISIGNRANK(...,'tail','neighbour','bound',[v1 v2]) computes the integral between v1 and v2 of the lower and upper distribution of P(Z>=-Z). Note that, 0<=v1=-Z) and [cred_bounds(2,1) cred_bounds(2,2)] is the credible interval of the upper distribution. Examples: x=randn(10,1); y=randn(10,1); [prob,h]=isignrank(y,x) x=randn(10,1); y=randn(10,1)+4; [prob,h]=isignrank(y,x) x=randn(10,1); y=randn(10,1); [prob,h,cred_bounds]=isignrank(y,x,'tail','both') See also ISIGNTEST, IRANKSUM. References:  A. Benavoli, F. Mangili, F. Ruggeri and M. Zaffalon "A Bayesian Wilcoxon signed-rank test based on the Dirichlet process" accepted to ICML 2014 (show less) Paper: http://ipg.idsia.ch/preprints/benavoli2014a.pdf Note that in the cloudrunner release of the software only some functionalities are actived. To use all the functionalities, download the Matlab or R sources from http://www.idsia.ch/~alessio/IDP.html Tags: test Wilcoxon signrank Bayesian Usage: Algorithm is public. Viewed 3416 times, called 114 times Upload:     1 vote Alessio Benavoli 05/15/2014 2:26 p.m. (version #13)

### Run Algorithm

 y: Y input vector x: X input vector alpha: Alpha level tail: Test type: 'right','left' s: Dirichlet Process prior strength nsamples: Number of Monte Carlo samples Cache: Allow cached result?
Close

### Please Wait

Computation is running...

### Plots

showing plot # of

### Resource Usage

Execution of the algorithm took seconds.
Peak memory usage for running the algorithm was kilobytes (total usage: kilobytes).

ID of this run:
Link:

### Result from Cache

Result has been delivered from cache and computed on (UTC).

(show)

### Error

Using this algorithm in your local MATLAB environment is easy: Click here for instructions!

### Usage Instructions for CloudRunner Client

1. Download the CloudRunner Client by clicking here and place the downloaded file in your MATLAB working directory.

2. Inside MATLAB, initialize the CloudRunner Client by calling CloudRunner:
>> CloudRunner

A login dialog will prompt for your CloudRunner mail address and password. For a start, you can leave the dialog empty and just click "Connect".

Alternatively, you can provide the login credentials (or empty strings to skip login) as a parameter and hence skip the login dialog. This is useful when using CloudRunner in non-interactive scripts.
>> CloudRunner('mail@example.com', 'password')

3. Select this algorithm by its URL. Selecting an algorithm creates the lambda function that proxies calls to the algorithm to the server for execution:
>> CloudRunnerSelect('http://www.cloudrunner.eu/algorithm/161/isignrank/version/13/')

For the sake of convenience, you can also use the algorithm ID instead of its URL for public algorithms.

4. Call functions from the algorithm like any regular local function.

Note: You can find further information on the help page.

### Comments

Please login to post a comment.

#1: Alessio Benavoli on 05/15/2014 2:41 p.m.
Note that in the cloudrunner release of the software only some functionalities are actived. To use all the functionalities, download the Matlab or R sources from
http://www.idsia.ch/~alessio/IDP.html