How to find the number of non-NaN elements in a column that are NaN in the last column, in MatLab?

Let's say I have a matrix

50 15 20 35 20

NaN NaN NaN 25 20

15 20 25 NaN NaN

NaN NaN 35 20 15

I want to find out the number of elements that are non-NaN in a column but are NaN in the last column.

For example, in Column 4, there is 1 element (4,2) that is non-NaN in Column 4 but is NaN in Column 3.

How do I find this number for each column?

Thank you so much!

Answer 1

[James Tursa]

Code is written from a slightly reordered wording:

"for each column, number of elements that are NaN in the last column and non-NaN in a column"

sum(isnan(YourMatrix(:,end)) & ~isnan(YourMatrix(:,1:end-1)))

E.g.,

YourMatrix = [
    50    15    20    35    20
    NaN NaN NaN 25    20
    15    20    25    NaN NaN
    NaN NaN 35    20    15]

YourMatrix = 4×5

50 15 20 35 20 NaN NaN NaN 25 20 15 20 25 NaN NaN NaN NaN 35 20 15

sum(isnan(YourMatrix(:,end)) & ~isnan(YourMatrix(:,1:end-1)))

ans = 1×4

1 1 1 0

Result is the number of elements matching criteria for each column except the last column.

[Julia]

Thank you so much for your answer!

May I ask a follow up question?

By "the last column", I meant "the column on the left" (eg. this column - Column 3, last column - Column 2). I tried to adjust your code like this, but the result is not what I want. Would you please tell me what I did wrong?

YourMatrix = [
    50    15    20    35    20
    NaN NaN NaN 25    20
    15    20    25    NaN NaN
    NaN NaN 35    20    15]

YourMatrix = 4×5

50 15 20 35 20 NaN NaN NaN 25 20 15 20 25 NaN NaN NaN NaN 35 20 15

for col = 1:5
sum(isnan(YourMatrix(:,1:col-1)) & ~isnan(YourMatrix(:,col)))
end

ans =

  1×0 empty double row vector

ans = 0

ans = 1×2

1 1

ans = 1×3

2 2 1

ans = 1×4

2 2 1 0

[James Tursa]

So, if I interpret "last column" to mean "previous column", maybe this is what you want?

YourMatrix = [
    50    15    20    35    20
    NaN NaN NaN 25    20
    15    20    25    NaN NaN
    NaN NaN 35    20    15]

YourMatrix = 4×5

50 15 20 35 20 NaN NaN NaN 25 20 15 20 25 NaN NaN NaN NaN 35 20 15

[0 sum(isnan(YourMatrix(:,1:end-1)) & ~isnan(YourMatrix(:,2:end)))]

ans = 1×5

0 0 1 1 0

If this isn't what you want, maybe you can show us what you expect the exact outcome to be, and then maybe we can figure out what you really want calculated.

[Walter Roberson]

YourMatrix = [
    50    15    20    35    20
    NaN NaN NaN 25    20
    15    20    25    NaN NaN
    NaN NaN 35    20    15];
sum(isnan(YourMatrix(:,1:end-1)) & ~isnan(YourMatrix(:,2:end)))

ans = 1×4

0 1 1 0

[Julia]

Yes, this is exactly what I want! Your answer is such a great help for me. Thank you so much for your patience!

Answer 2

[Walter Roberson]

isnan(YourMatrix(:, 3)) & ~isnan(YourMatrix(:, 4))