Splitting Table at NaN

I have 2 tables and I want to split them at NaN values, for now I am ignoring repeated NaNs but I want to consider them in the next step. What I have now :

x = table2array(t(:,1));
y= table2array(t(:,2));
index=find(~isnan(x));
idx=find(diff(index)~=1);
A=[idx(1);diff(idx);numel(index)-idx(end)];
x_points=mat2cell(x(~isnan(x)),A,1);
y_points=mat2cell(y(~isnan(y)),A,1);
    

If two NaNs occur at once, I want it to become a NaN cell.

Answer 1

[Vatsal]

Hi @Umair Shahid,

I understand that you are splitting the array “x” and ‘y” at the values of NaN. According to your code you are splitting “x” array at the values of NaN and wherever split happens in “x”, you are splitting “y” at the same location. Now you wanted to consider the repeated NaNs, which means if more than one consecutive NaNs are there you wanted to create a NaN cell also.

Below is the code which will consider repeating NaNs and provide the output as needed:

count = 0;
nanIndex = [];

nanIndices = find(isnan(x));

for i = 1:length(nanIndices)
    if i > 1 && nanIndices(i) == nanIndices(i-1) + 1
        count = count + 1;
    else
        count = 1;
    end
    
    if count > 2
        nanIndex(end +1) = nanIndices(i);
    end
end
x(nanIndex) = [];

disp(x);

nanIndices = find(isnan(x));

x_points = {};
y_points = {};

startIndex = 1;
for i = 1:length(nanIndices)
    endIndex = nanIndices(i) - 1;
    x_points{i} = x(startIndex:endIndex);
    y_points{i} = y(startIndex:endIndex);
    startIndex = nanIndices(i) + 1;
    
    if i > 1 && nanIndices(i) - nanIndices(i-1) > 1
        x_points{i-1} = NaN;
        y_points{i-1} = NaN;
    end
end

if startIndex <= length(x)
    x_points{end+1} = x(startIndex:end);
    y_points{end+1} = y(startIndex:end);
end

for i = 1:numel(x_points)
    if all(isnan(x_points{i}))
        x_points{i} = NaN;
        y_points{i} = NaN;
    end
end

I hope this helps!