How do I delete a group of rows when a condition is met in one?

Let's say I have a matrix which is 1000x1. When the value of a row > 0.25, I want to delete that row and 10 rows above and below. (e.g., of value of row 30 = 0.5, rwos 20-40 would be deleted). I have the follwoing starting point but not sure how to select the rows above and below.

data(data > 0.25,:) = []

Answer 1

[David Hill]

f=find(data>.25);
idx=[];
for k=1:length(f)
    idx=[idx,(f(k)-10):(f(k)+10)];
end
idx=unique(idx);
idx=idx(idx>0&idx<=length(data));
data(idx)=[];

Answer 2

[Fifteen12]

Simple but long answer using an iterator:

Edits: Changed code to catch moving indices as data is resized and expand removal range if trigger values are found in the initial set.

data1 = randi(10,10,1);
threshold = 7;
clear_range = 2; %rows/samples

data = clearRange(data1, threshold, clear_range);

function [data] = clearRange(data, threshold, clear_range, compare_column)
    % data: matrix of values to clean
    % threshold: delete values greater than threshold
    % clear_range: number of rows to delete
    % compare_column: column to check for value to compare against threshold
    if nargin < 4
        compare_column = 1;
    end

    i = 1;
    while i <= length(data)
        if data(i,compare_column) > threshold
            to_delete = [max(1, i-clear_range), min(length(data), i+clear_range)];
            j = i + 1;
            while j <= to_delete(2)
                if data(j, compare_column) > threshold
                    to_delete = [to_delete(1), min(j + clear_range, length(data))];
                end
                j = j + 1;
            end
            data(to_delete(1):to_delete(2), :) = [];
            i = max(1, i - clear_range);
        else
            i = i + 1;
        end
    end
end

[Benjamin Colbert]

This code is quick but unfortately it's not working quite right. When run, it leaves many high amplitude events which are over the threshold. Changing the clearing range helps with that but is finicky and doesn't fix it. In many cases multiple high amplitude values are directly next to each other (sometimes hundreds in a row) and I wonder if that's the issue. To give you a better idea of what I'm doing - I'm reading in aidio data [data,y] and which ahs 32768 values a second. There are more than a million values. These events are generall 8000 samples long (changing the theshold to 8000 or even 16000 samples doesn't help).

[Benjamin Colbert]

Here's a reproducable example, You can see it does not remove all values above the threshhold.

%[data, fs] = audioread("noise.wav");
%t1 = linspace(0, (numel(data)-1)/fs, numel(data));
rng(1)
data = randi(10,1000,1);
threshold = 5;
clear_range = 10; %rows/samples
data = clearRange(data, threshold, clear_range);
%t1 = linspace(0, (numel(data)-1)/fs, numel(data));
%plot(t1, data);
plot(data)
function [data] = clearRange(data, threshold, clear_range, compare_column)
% data: matrix of values to clean
% threshold: value to compare values against
% clear_range: number of rows to delete
% compare_column: column to check for value to compare against threshold
if nargin < 4
    compare_column = 1;
end
for i = 1:length(data)
    if i > length(data)
        break
    end
    if data(i,compare_column) > threshold
        data(max(1, i-clear_range):min(length(data), i+clear_range),:) = [];
    end
end
end

[Fifteen12]

Good catch. The problem was that the data matrix was being shifted every time a range was cleared, so the for loop was essentially jumping "clear_range" rows every time it was called.

There was an additional item of consideration where values over the threshold that are within the clearing range of another value are deleted before they can be considered by the algorithm. I've also edited the code to check each value before it's deleted, and extend the clearing range if a trigger value is found.

The code is pretty unoptimized, again, simple but slow in this case. You'd get much better performance by vectorizing. @David Hill's solution works fine and solves both of these problems with much better performance. I can't think of any reason to utilize mine versus his.