# Hi everyone, I have to use monte carlo time series method to generate synthetic smart meter data for about a million houses. Any heads up on this will be really helpful. Thanks

11 views (last 30 days)
Ragini Gupta on 29 Oct 2017
Edited: Walter Roberson on 4 Nov 2017
Monte Carlo Time Series Synthetic Data generation

Image Analyst on 30 Oct 2017
Edited: Image Analyst on 30 Oct 2017
Search for Monte Carlo for lots of posts, many of which have code. https://www.mathworks.com/matlabcentral/answers/?term=tag%3A%22Monte+Carlo%22
I've attached some demos that use Monte Carlo.
Ragini Gupta on 4 Nov 2017
Edited: Walter Roberson on 4 Nov 2017
Thank you so much. Turns out I have to model the data using Monte Carlo Markov Chain. I created the Transition probability matrix from the consumption values after the binning stage. However, I am not sure how to proceed to simulate the matrix and generate new consumption values.
Here is the code, if you can have a look at it? Please. Thanks in advance
filename = 'newTestingExcel.xlsx';
edges = linspace(min(Furnace),max(Furnace),8);
[counts,bins] = histc(Furnace, edges);
[counts,bins] = histc(Furnace, edges);
%# transition matrix
m = max(bins);
n = numel(bins);
y = zeros(m,1);
p = zeros(m,m);
for k=1:n-1
y(bins(k)) = y(bins(k)) + 1;
p(bins(k),bins(k+1)) = p(bins(k),bins(k+1)) + 1;
end
p = bsxfun(@rdivide,p,y); p(isnan(p)) = 0;
disp(trans);