Is there a commonly accepted practice for calculating (and/or visualizing) correlation between angular variables? (I.e. 359° is pretty close to 1°.) I'm sure I could concoct a suitable trigonometric transformation, but would rather stick to convention if one exists.
Asked
Active
Viewed 156 times
2
-
1Possible duplicate of [regression for angular/circular data](http://stats.stackexchange.com/questions/109959/regression-for-angular-circular-data) – rep_ho Oct 17 '16 at 12:19