Data set must contain at least 3 samples error in find peaks embedded in for loop

Question

Christiane Voufo 2019-11-10

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/490203-data-set-must-contain-at-least-3-samples-error-in-find-peaks-embedded-in-for-loop

评论： Star Strider 2019-11-10

Hi All,

I am trying to iterate through the columns of a matrix I have called 'data' to find the peaks of each column. To do this, I have a for loop that essentially looks like this:

for i = 1:size(data,2);

[pks{i,:},locs{i,:}] = findpeaks(bcgFixed(i,:),'MinPeakProminence',(bcgthresh*bcgStd),'MinPeakDistance', ((6/frate)* 10^-3)); %find all points where the baseline-subtracted background trace crosses 2*std

numEvents = length(locs)

% disp(locs)

% figure

% findpeaks(bcgFixed(i,:),'MinPeakProminence',(bcgthresh*bcgStd),'MinPeakDistance', ((1/frate)* 10^-3));

end

when I try to run the code it gives me the error message in the title. I'm not sure why because this matrix contains a lot more than 3 samples. Would really appreciate some help with this one.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Star Strider 2019-11-10

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/490203-data-set-must-contain-at-least-3-samples-error-in-find-peaks-embedded-in-for-loop#answer_400680

在 MATLAB Online 中打开

‘... columns of a matrix ...’

Note that:

[pks{i,:},locs{i,:}] = findpeaks(bcgFixed(i,:),'MinPeakProminence',(bcgthresh*bcgStd),'MinPeakDistance', ((6/frate)* 10^-3)); %find all points where the baseline-subtracted background trace crosses 2*std

will iterate across rows.

To iterate across columns, use:

bcgFixed(:,i)

instead:

[pks{i,:},locs{i,:}] = findpeaks(bcgFixed(:,i),'MinPeakProminence',(bcgthresh*bcgStd),'MinPeakDistance', ((6/frate)*1E-3)); %find all points where the baseline-subtracted background trace crosses 2*std

The cell arrays storing the output do not care, however the input array does.

12 个评论
显示 10更早的评论隐藏 10更早的评论

Christiane Voufo 2019-11-10

在 MATLAB Online 中打开

Okay, here is the full code:

function numEventsPerArea(AllCellsFOV1910150007)
data = AllCellsFOV1910150007;
%takes array [m,n] where m is time (or frame) and n is ROI number; values
%are intensities at each timepoint. Assumes that first trace is full field
%dFoF trace.
roithresh= 0.5;
Fs=2.96;
frate=Fs;
frames=1750;
bcgthresh= 1;
valarray = AllCellsFOV1910150007;
numel(valarray(:,:));
for i = 1:size(valarray(:,:))
    for i = 1:size(valarray,2)
        ff(i) = valarray(:,i)%get neuronal trace
    end
    for i = valarray(:,1)
        bcgMean(i) = mean(ff(:,i));    %find mean
    end
    for i = valarray(:,1)
        bcgFixed(i) = (ff(:,i) - bcgMean(:,i)); %subtract mean to center at zero
    end
    %find the standard deviation of the trace
    bcgStd(i) = std(bcgFixed(:,i));
end
for i= size(bcgFixed);
    [pks{i},locs{i}] = findpeaks(bcgFixed(:,i),'MinPeakProminence',(bcgthresh*bcgStd),'MinPeakDistance', ((6/frate)* 10^-3)); %find all points where the baseline-subtracted background trace crosses 2*std
    numEvents = length(locs)
%     disp(locs)
%     figure
%     findpeaks(bcgFixed(i,:),'MinPeakProminence',(bcgthresh*bcgStd),'MinPeakDistance', ((1/frate)* 10^-3));
end 
end

What I was trying to do was get bcgFixed to be a 1750x577 array, by adding the extra for loops in the beginning, but it got really convoluded and I'm not sure how to make it run smootly.

Christiane Voufo 2019-11-10

在 MATLAB Online 中打开

Could I get help on this last one? The error says that the matrices I'm trying to subtract are not equal. They both have the same size, but column 1 in matrix bcgMean is all zeros and I think this is the problem, althouhg I'm not sure why the first row of this matrix is not getting populated in the first for loop. I'm not sure if this is even the problem.

bcgMean = zeros(1,577);
bcgMean1 = zeros(1750,577);
bcgFixed = zeros(1750,577);
for i2 = 2:size(valarray,2)
    bcgMean(i2) = mean(valarray(:,i2));%find mean
    bcgMean1 = repelem(bcgMean(1:1750:end,:),1750,1);
end
for i = 2:size(valarray,2)
    bcgFixed(i) = (valarray(:,i) - bcgMean1(:,i2)); %subtract mean to center at zero
end

Star Strider 2019-11-10

在 MATLAB Online 中打开

I apologise for the delay — sleeping.

I cannot understand what you are doing, especially with:

bcgMean1 = repelem(bcgMean(1:1750:end,:),1750,1);

Note that you can calculate the row means of ‘valarray’ as:

bcgMean = mean(valarray);

so including it in the loop is unnecessary.

Here you are using the last column (whatever the last value of ‘i2’ was in the previous loop) of ‘bcgMean1’:

 bcgFixed(i) = (valarray(:,i) - bcgMean1(:,i2)); %subtract mean to center at zero

from all of ‘valarray’, column-wise. If they do not have the same row sizes, that would throw the error you are seeing.

请先登录，再进行评论。

Data set must contain at least 3 samples error in find peaks embedded in for loop

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

采纳的回答

12 个评论
显示 10更早的评论隐藏 10更早的评论

更多回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

Data set must contain at least 3 samples error in find peaks embedded in for loop

0 个评论 显示 -2更早的评论隐藏 -2更早的评论

采纳的回答

12 个评论 显示 10更早的评论隐藏 10更早的评论

更多回答（0 个）

另请参阅

类别

标签

Community Treasure Hunt

WeChat

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

12 个评论
显示 10更早的评论隐藏 10更早的评论