MATLAB: Ranksum requires tiedrank

MATLABranksumStatistics and Machine Learning Toolboxtiedrank

I assumed that the function ranksum computed ranks, etc. automatically, but I ran across the documentation for tiedrank and now I am not sure. Do I need to run my data through tiedrank before running it through ranksum?

I am using ranksum because my data are non-normal, not because the data are inherently ordinal.

Best Answer

No, you do not. ranksum() calls tiedrank() internally. You can see this by typing

 >> edit ranksum

Related Solutions

MATLAB: Are there functions for calculating the PDF and CDF of the Pareto distribution in the Statistics Toolbox

This enhancement has been incorporated in Release 14 Service Pack 3 (R14SP3). For previous product releases, the Statistics Toolbox has no functionality for computing the PDF or CDF of the Pareto distribution.

MATLAB: Tiedrank in descending order

Hi Ekin,

You can the reverse the order of ranking programmatically as follows:

 %Store the result of tiredrank function in "a"
a = tiedrank([10 20 30 40 20]);
%Find the largest rank value of "a", increment it by 1 and subtract "a" by
%this value.
ranks_desc = (max(a)+1) - a;

This will just reverse the ranking received by "tiedrank" function.

Best Answer

Related Solutions

MATLAB: Are there functions for calculating the PDF and CDF of the Pareto distribution in the Statistics Toolbox

MATLAB: Tiedrank in descending order

Related Question