corr(x,y) nan

Hello, I am trying to run a correlation using corr(x,y), but my dataset has NaNs in it, for different participants (beginner coder here).

For example, my dataset and code may look like this:

a = [78 88 87 54 68 91 72 nan 88 nan];
b = [nan 35 42 44 67 55 nan 34 25 33];
x = ~isnan(a);
y = ~isnan(b);
[r,p] = corr(x,y)

The result is a table full of NaN in the ouptut.

I am trying to get the r value and the p value. Would there be another way to achieve this?

Answer 1

[Michael]

I think you are looking for corrcoef instead of corr. To exclude the nan values, use

a = [78 88 87 54 68 91 72 nan 88 nan];
b = [nan 35 42 44 67 55 nan 34 25 33];
r = corrcoef(a,b, 'rows','complete')

[madhan ravi]

You forgot to provide the function in the last line.

[Michael]

Ups, I'm sorry. I edited my posting.