I have to generate transition probability matrix from sample data.

I have the following code of generating transition probability matrix.

data=randi(12,[5,3]);

%%First bin data into categories

speedBinN = 5;

aceelBinN = 5;

speed = histc( data(:,2), linspace(min(data(:,2)),max(data(:,2)),speedBinN) ); % bin them into categories

accel = histc( data(:,3), linspace(min(data(:,3)),max(data(:,3)),aceelBinN) );

%%count up transitions

transCountMat = zeros(speedBinN,aceelBinN,speedBinN,aceelBinN);

for ii = 1:size(data,1)-1

transCountMat( speed(ii),accel(ii),speed(ii+1),accel(ii+1) ) = transCountMat( speed(ii),accel(ii),speed(ii+1),accel(ii+1) ) + 1;

end

%%calculate probabilities

sumOverPossibleDestinations = sum( sum(transCountMat, 4), 3);

transMat = bsxfun( @rdivide, transCountMat, sumOverPossibleDestinations );

%%User Interactive stuff

IM = imagesc(squeeze(transMat(1,1,:,:)));

colorbar

set(IM,'ButtonDownFcn',@bdFcn)

set(gca,'ydir','normal')

ylabel speed

xlabel accel

hold on

p = plot(1,1,'w');

updateIndicator(1,1)

but it generates the following error:

Attempted to access transCountMat(2,1,0,1); index must be a positive integer or logical.

Error in Untitled2 (line 12)

transCountMat( speed(ii),accel(ii),speed(ii+1),accel(ii+1) ) = transCountMat(

speed(ii),accel(ii),speed(ii+1),accel(ii+1) ) + 1;

Plz resolve it.

Answer 1

[Shaunak]

Hi Sami,

It is my understanding that you're trying to generate a transition probability matrix in MATLAB, but you're encountering an error related to indexing:

"Attempted to access transCountMat(2,1,0,1); index must be a positive integer or logical."

This issue arises because one of your indices is zero, which is not allowed in MATLAB.

You can use the ‘discretize’ function instead of ‘histc’ to ensure that your indices are valid positive integers. This will help avoid the zero-indexing issue.

Here’s an alternative way for your reference:

data = randi(12, [5, 3]);
% First bin data into categories
speedBinN = 5;
accelBinN = 5;

% Define bin edges
speedEdges = linspace(min(data(:,2)), max(data(:,2)), speedBinN+1);
accelEdges = linspace(min(data(:,3)), max(data(:,3)), accelBinN+1);

% Bin the data
speed = discretize(data(:,2), speedEdges);
accel = discretize(data(:,3), accelEdges);

% Count up transitions
transCountMat = zeros(speedBinN, accelBinN, speedBinN, accelBinN);
for ii = 1:size(data, 1) - 1
    if speed(ii) > 0 && accel(ii) > 0 && speed(ii+1) > 0 && accel(ii+1) > 0
        transCountMat(speed(ii), accel(ii), speed(ii+1), accel(ii+1)) = ...
            transCountMat(speed(ii), accel(ii), speed(ii+1), accel(ii+1)) + 1;
    end
end

% Calculate probabilities
sumOverPossibleDestinations = sum(sum(transCountMat, 4), 3);
transMat = bsxfun(@rdivide, transCountMat, sumOverPossibleDestinations);

% User Interactive stuff
IM = imagesc(squeeze(transMat(1, 1, :, :)));
colorbar
set(IM, 'ButtonDownFcn', @bdFcn)
set(gca, 'ydir', 'normal')
ylabel('speed')
xlabel('accel')
hold on
p = plot(1, 1, 'w');
updateIndicator(1, 1) % Assuming the function updateIndicator is defined

You can find more information on the ‘discretize’ function in the following MATLAB documentation: https://www.mathworks.com/help/matlab/ref/double.discretize.html

Hope this helps!