Using accumarry and changing size of output

Question

susman 2021-2-21

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/752089-using-accumarry-and-changing-size-of-output

评论： susman 2021-2-22

I am using the accumarray by using the following command on the first and second column of struct array. The first part of the code works good.

but the last two commands "went_from_to_countAge" and "went_from_to_probAge", does not work. any idea?

sequence = [1 1 1 1 2 2 2 2 2 3;...
            1 1 1 1 1 1 1 1 1 1;...
            1 1 1 1 1 1 1 1 1 1;...
            1 1 1 1 1 1 1 2 2 2;...
            1 1 1 1 1 1 1 1 1 1;...
            2 2 2 2 2 2 2 2 2 2; ...
            2 2 2 2 2 2 2 3 3 3; ...
            2 3 3 3 3 3 2 2 2 2; ...
            2 2 2 2 2 2 4 4 4 4]
% first part of the code 
k = 1:size(sequence)
Age (k) = 25 + k
States = unique([sequence(:,k); sequence(:,k+1)]); % Find the unique rows based on the data in the first two columns. 
[TF, fromstateAge] = ismember(sequence(:,k), States); % logical arrays 
[TF, tostateAge] = ismember(sequence(:,k+1), States);
% second part of the code 
went_from_to_countAge(:,k) = accumarray( [fromstateAge(:), tostateAge(:)], 1, []) % frequency of each unique number 
went_from_to_probAge(k)= went_from_to_countAge ./ sum(went_from_to_countAge); % pribability of each unique number 

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

dpb 2021-2-21

Attach a .mat file with a sample of the struct and an example of what the desired output should be.

Almost impossible to visualize such complexity without having an example at hand.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Walter Roberson 2021-2-21

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/752089-using-accumarry-and-changing-size-of-output#answer_629494

在 MATLAB Online 中打开

I am not clear if you are saying that you want different sizes to be returned? If so then you either need to use a cell array or else you have to pack the results into a larger array with padding.

If instead the point is that the size returned is sort of by chance but you want a fixed size, then change the place you used [] to give the required output size, such as

A = accumarray( [data(1).fromstate(:,2), data(1).tostate(:,3)], 1, [5 5]);

7 个评论
显示 5更早的评论隐藏 5更早的评论

susman 2021-2-22

在 MATLAB Online 中打开

sequence = [1 1 1 1 2 2 2 2 2 3;...
            1 1 1 1 1 1 1 1 1 1;...
            1 1 1 1 1 1 1 1 1 1;...
            1 1 1 1 1 1 1 2 2 2;...
            1 1 1 1 1 1 1 1 1 1;...
            2 2 2 2 2 2 2 2 2 2; ...
            2 2 2 2 2 2 2 3 3 3; ...
            2 3 3 3 3 3 2 2 2 2; ...
            2 2 2 2 2 2 4 4 4 4]
States = unique([sequence(:,1); sequence(:,2)]); % Find the unique rows of A based on the data in the first two columns. 
[TF, fromstate] = ismember(sequence(:,1), States); % logical arrays 
[TF, tostateAge] = ismember(sequence(:,2), States);
went_from_to_count_age_25 = accumarray( [fromstate(:), tostateAge(:)], 1, []); % frequency of transitions from one state to another 
s = sum(went_from_to_count_age_25, 2);
went_from_to_prob_age_25 = went_from_to_count_age_25 ./ s;

For example, for the above code, I calculate "went_from_to_prob_age_25" using the column 1 and column 2 of the matrix "sequence".

How can I make a loop, such that I can do that for

"went_from_to_prob_age_26" using column 2 and 3

"went_from_to_prob_age_27" using column 3 and 4

.....

susman 2021-2-22

在 MATLAB Online 中打开

and this remains as it is?

went_from_to_countAge(:,k) = accumarray( [fromstateAge(:), tostateAge(:)], 1, []) % frequency of each unique number 
went_from_to_probAge(k)= went_from_to_countAge ./ sum(went_from_to_countAge); % pribability of each unique number

Sorry for asking these stupid questions, as I already spend enough time to figure it out.

susman 2021-2-22