RemoveValuesUnderHistogramTreshold

RemoveValuesUnderHistogramTreshold.m

function [ data_filtered ] = RemoveValuesUnderHistogramTreshold( data, percentageTreshold, numberOfHistogramBins )
%RemoveValuesUnderHistogramTreshold This function removes outstanding data variables that is
%   under specific occurences threshold. The occurrences is partitioned
%   by histogram method, equals bin size and fixes number of bins.
%
%   data_filtered = RemoveValuesUnderHistogramTreshold(DATA, percentageTreshold, numberOfHistogramBins)
%       (matrix) DATA : The vector with data
%       (double) percentageTreshold  : The occurences percentage (relative to maximum)
%       that rejects outstanding values. Double value, e. g. 0.5 for under
%       50%
%       (integer) numberOfHistogramBins : Number of hist() bin counts
%   
%   Author: Piotr Pelczar ([email protected])
%
    edges_min = min(data);
    edges_max = max(data);
    edges_step = abs((edges_max-edges_min)/(numberOfHistogramBins - 1));
    
    edges = edges_min:edges_step:edges_max;
    [bincount, data_bin_idx] = histc(data, edges);
    
    bincount_max = max(bincount);
    
    edges_under_treshold = ((bincount/bincount_max) < percentageTreshold);
    edges_under_treshold_idx = 1:10;
    edges_under_treshold_idx = edges_under_treshold_idx(edges_under_treshold);
    data_bin_idx_under_treshold = ismember(data_bin_idx, edges_under_treshold_idx);
    
    data_filtered = data;
    data_filtered(data_bin_idx_under_treshold) = [];
end