How do I using logical values to write 'NAN' to non-existing data?

I have a data set with non equal lengths

A = B =

H m lat H m lat

0 2 -90 0 1 -90

0 2 -89 0 1 -89

0 2 -88 0 1 -87

0 2 -87 0 1 -86

0 2 -86

I want to query using the lat of A to find missing lat in B and write 'NAN' in that row

so that the sizes become the same like this; I want to use this idea for a large data set

B=

0 2 -90

0 2 -89

NAN NAN NAN

0 2 -87

0 2 -86

Answer 1

[Rik]

You can use ismember:

A=[ones(5,2).*[0 2] (-90:-86).'];
B=[ones(4,2).*[0 2] [-90;-89;-87;-86]];

new_B=nan(size(A));
[~,rows]=ismember(B(:,3),A(:,3));
new_B(rows,:)=B;
new_B

new_B = 5×3

0 2 -90 0 2 -89 NaN NaN NaN 0 2 -87 0 2 -86

[Stephen Tete]

thank you. it worked

[Stephen Tete]

Now i want to automate it for cells (the cell is up to 200) each containing data

i did use the for loop but it only returns the last dataset see here

refom = nan(size(A{1}));
for i = 2 : length(A)
    [~,rows] = ismember(A{i}(:,3),A{1}(:,3));
    refom(rows,:) = A{i};
end

Can you help me figure it out please

[Rik]

You have posted code, but you didn't explain how the output is different from what you want.

Note that this code will overwrite any values that occur multiple times in later elements of A.

[Stephen Tete]

A is a cell like this

A = [180x4 double; 179x4 double; 180x4 double; 180x4 double; 178x4 double; 177x4 double; ...]

class(A)
>> 'cell'

each cell contains dataset with first element in A having full data set (size 180x4). some of the proceeding data do not have the full size (ex. 178x4 and 177x4). In this case i want to use column 3 of the first cell data to query those that do not have the full data(all data has the same elements in column 3) and put NAN at rows where there is no data.but i do not want to overwrite the outputs.

When i use your initial code to do for only two cells it works nicely but now i want to loop through all the cells in A and do for each of them once and save the output cells having all 180x4 after inserted inserting NANs

That was why i used this code

refom = nan(size(A{1}));
for i = 2 : length(A)
    [~,rows] = ismember(A{i}(:,3),A{1}(:,3));
    refom(rows,:) = A{i};
end

the output i desire should look like this

refom = 
180x4 double
180x4 double
180x4 double
180x4 double
180x4 double
180x4 double

% so that if i open each cell i see data having both int and nans

but i get only the last cell which means the code overwrites the preceeding outputs to display the final last one

i = 200

refom = 180x4

-------------------------------------------------------------------------------------------

you can find few example of data in each cells here with heads

%   H   M   lat     TC       
    0	2	-90	0.701661800000000
    0	2	-89	1.38124329032258
    0	2	-88	1.50675797468354
    0	2	-87	0.934005460000000
    0	2	-86	1.30263100000000
    0	2	-85	1.42799615000000
    0	2	-84	1.94040264556962
    0	2	-83	1.67793006666667

I hope this is elaborate