2.13.4.5 mwtest(Pro)

1 Menu Information
2 Brief Information
3 Additional Information
4 Command Line Usage
5 X-Function Execution Options
6 Variables
7 Examples
8 More Information
9 Related X-Functions

Menu Information

Statistics: Nonparametric Tests: Mann-Whitney test

Brief Information

Preform Mann-Whitney test

Additional Information

This feature is for OriginPro only.

Command Line Usage

1. mwtest  type:=1 irng:=(1,2) 
2. mwtest  irng:=(1,2) tail:=upper; 
3. mwtest  irng:=(1,2) tail:=lower rt:=<new MW>

X-Function Execution Options

Please refer to the page for additional option switches when accessing the x-function from script

Variables

Display
Name

Variable
Name

I/O
and
Type

Default
Value

Description

Input Data Form

type

Input

int

Specify the input data form.

Option list:

Indexed
Input data range, including input group range and data range.

Example:

mediantest (1,2)

In this example, the input group should be stored in the first column and the input data should be stored in the second column.

Raw
Specify two data columns as input data.

Note: With the Stack and Unstack tool, you can convert indexed data to raw data, and raw data to indexed data.

Input

irng

Input

Range

Specify the input data.

Null Hypothesis

null

Input

string

Specify the null hypothesis of Mann-Whiney test.

Alternate Hypothesis

tail

Input

int

Radio

Indicates whether an upper, lower, or two-tailed MW-test should be performed.

Option list:

0=Radio:two:Median1 <> Median2

Performs a 2-tailed MW-test

1=upper:Median1 > Median2

Performs an upper-tailed MW-test

2=lower:Median1 < Median2

Performs an lower-tailed MW-test

Significance Level

alpha

Input

double

0.05

Specify the significance level of the test.

Exact P Value

exact

Input

int

Specify whether to calculate exact p value.

Output Results

Output

ReportTree

[<input>]<new>

Include Notes table, Descriptive table, Ranks table, and Test Statistics table.

Examples

To perform a two-tailed MW-test on columns 1 and 2 of the active worksheet, using default settings:

mwtest irng:=(1,2)

To perform a lower-tailed MW-test on columns 1 and 3 of the active worksheet, and save the result table with the name MW, use the script command:

mwtest irng:=(1,3) tail:=2 rt:=<new name:=MW>

Sample Code

/*
This example is used to show how to test whether the two populations have identical distribution or not,
when the normality is questionable.
The sample data used is in
  OriginPath\Samples\Statistics folder.
1. Import the sample data into a book in Origin
2. use mwtest XF to calculate the U-statistics and p-value.
3. put the result into a new sheet
*/
/*Import the sample data into a new book*/
String fname$=system.path.program$+"Samples\Statistics\mw-test.dat";
newbook;
impASC;
string bkn$=%H;

/*Use the mwtest XF to calculate the U statistics and p-value*/
mwtest irng:=[bkn$]1!(col(1),col(2)) rt:=<new name:="Manny-Whitney">;

/*New a sheet to stor the results of the Manny-Whitney test*/
newsheet book:=bkn$ name:="Result" label:="MW_U|Zstat|Sig";
range MW_U=[bkn$]Result!col(1);
range Zstat=[bkn$]Result!col(2);
range Sig=[bkn$]Result!col(3);
getresults iw:=[bkn$]2 tr:=mytree;
MW_U[1]=mytree.stats.stats.c1;
Zstat[1]=mytree.stats.stats.c2;
Sig[1]=mytree.stats.stats.c3;