# Random picking of adjacent data points

3 views (last 30 days)
aa on 29 Aug 2020
Commented: aa on 30 Aug 2020
Hi,
May someone help me
I have data in one column and want to randomly pick adjacent 48 data points for 1000 times. For example, we have 10 data point (1, 2,3, 4,5,6,7,8,9,10) and we randomly pick 3 adjacent data points 5 times.
May some one help me ...
Thank you. Image Analyst on 29 Aug 2020
Try this:
data = [1:50000]'; % Whatever.....
% Define general parameters for getting the samples:
numElements = size(data, 1)
numSubsets = 1000
% Now do the iterations to get each of the 1000 subsets.
for k = 1 : numSubsets
% Get the first index at random. Don't get within numAdjacent of the end though!
firstIndex = randi(numElements - numAdjacent + 1);
% Now get the second index which will make the subset have adjacent indexes.
secondIndex = firstIndex + numAdjacent - 1;
% Get one subset of 48 adjacent data samples.
theseData = data(firstIndex : secondIndex);
% Now do something with this set of adjacent elements.
end

aa on 29 Aug 2020
Thnk you for explaining this, Yes, code works well there is no error but it did not clculate for the whole loop. This gives just one column of 48 data point but i required 1000 columns.
Image Analyst on 29 Aug 2020
You said "I have data set of 50000 enteries and I want to pick 1000 random subsets having 48 adjacent enteries." You did not tell us that you wanted those subsets stored in a 2-D array. Well I'm sure you've already done it by now, but anyway, here is how I created a matrix of 1000 columns and put each subset into its appropriate column:
data = [1:50000]'; % Whatever.....
% Define general parameters for getting the samples:
numElements = size(data, 1)
numSubsets = 1000
% Create an output array with 1000 columns.
% Now do the iterations to get each of the 1000 subsets.
for k = 1 : numSubsets
% Get the first index at random. Don't get within numAdjacent of the end!
% Now get the second index which will make the subset adjacent.
secondIndex = firstIndex + numAdjacent - 1;
fprintf('Subset %d goes from index %d to index %d.\n', k, firstIndex, secondIndex);
% Get one subset of 48 adjacent data samples.
theseData = data(firstIndex : secondIndex);
% Now store this set of adjacent elements into our output array.
output(:, k) = theseData;
end
aa on 30 Aug 2020
Thank you, THis works very well and serve the purpose perfectly.

madhan ravi on 29 Aug 2020
Edited: madhan ravi on 29 Aug 2020
datasets = num2cell(reshape(data(randperm(48e3)), [], 1e3), 1);
%celldisp(datasets)

aa on 29 Aug 2020
thanks for this, but this not work in my case, i want to get adjacent enteries. For example, I have data set of 50000 enteries and I want to pick 1000 random subsets having 48 adjacent enteries.
Hope this i conveny you exactly.
madhan ravi on 29 Aug 2020
Edit your question a billion times to each new question.
aa on 29 Aug 2020
Apology, I may not explain what i exactly want ... This may explain my question in better way ..... 