Why is corr/corrcoef returning NaNs?

I have 2 vectors that I would like to correlate with one another: OptimismBias and AmbigTolG (listing the first 10 values out of 248).

-0.0263 0.0000

0.7971 0.0000

-0.2800 -0.0000

0.4088 0.0000

0.6691 0.0000

0.5348 NaN

0.2606 0.0000

0.6995 NaN

-0.3270 0.0000

0.4665 0.0139

When I run the corr, I get all NaNs at the output. I'm aware that the AmbigTolG values are 0 or close to 0 but I'm not sure how else to run this.

[Tommy]

What syntax are you using? Perhaps

corr(OptimismBias, AmbigTolG, 'rows', 'complete')

to ignore the NaN values?

[Darla Bonagura]

That solved it. Thank you!

[Adam Danz]

Tommy's comment should be an answer.

[Adam Danz]

Related to this topic: Getting a NaN in correlation coefficient.

Answer 1

[Tommy]

You can use

corr(OptimismBias, AmbigTolG, 'rows', 'complete')

to ignore the NaN values.