MATLAB: Is the “prctile” function returning the same output for different input

MATLABprctile

I have a vector of data called "var", and I noticed that the two function calls to "prctile":

prctile(var(1 : end), 0.008)

and

prctile(var(1 : end), 0.016)

return the same result. Is this an error or bug?

Best Answer

This is intended behavior.

For percentile values outside the range of data, "prctile" will output the minimum or maximum of the data as appropriate. In this case the "prctile" function was returning the minimum value in the data both times, because both .008 and .016 are outside the minimum percentile.

For more information about the algorithms involved in the "prctile" function, please view the following documentation, under the "Algorithms" section:

http://www.mathworks.com/help/stats/prctile.html

Related Solutions

MATLAB: Is the Score Transform lost when using “saveCompactModel”

One of the limitations of using "saveCompactModel" to save an SVM model with capabilities to predict posterior probabilities is that "loadCompactModel" cannot restore the 'ScoreTransform' property to the MATLAB Workspace.

However, while preforming code generation, "loadCompactModel" is able to load the model along with the 'ScoreTransform' property at compile time within an entry function, and hence should function properly.

For more information on such limitations on various model objects while using "loadCompactModel", please refer to the documentation link below:

https://www.mathworks.com/help/stats/loadcompactmodel.html#bvcl05n-1-CompactMdl

MATLAB: Obtaining a list of percentages/cumulative probabilities from a data vector given the percentile/quantile values

While MATLAB does not include an inbuilt function to accomplish this, it is possible to write some simple code to do so. The following page of the MATLAB documentation provides details of the algorithm used by the "prctile" and "quantile" functions.

http://www.mathworks.com/help/stats/quantiles-and-percentiles.html

The following code can be used to reverse these algorithms to arrive at the percentage and cumulative probability values given a data vector "X" and a list of percentile or quantile values "Y".

For "prctile",

>> sortedX = sort(X);
>> listPercentages = (100/length(X))*(0.5:1:length(X)-0.5);
>> percentages = interp1(sortedX,listPercentages,Y);

For "quantile",

>> sortedX = sort(X);
>> listCumProb = (1/length(X))*(0.5:1:length(X)-0.5);
>> cumProbs = interp1(sortedX,listCumProb,Y);

Best Answer

Related Solutions

MATLAB: Is the Score Transform lost when using “saveCompactModel”

MATLAB: Obtaining a list of percentages/cumulative probabilities from a data vector given the percentile/quantile values

Related Question