I need to understand how to read kernel density graphs. How do you come up with the values in y-axis?
$\begingroup$
$\endgroup$
1
-
$\begingroup$ Also closely related: stats.stackexchange.com/questions/4220/… and stats.stackexchange.com/questions/14483/…. $\endgroup$– whuber ♦Commented Apr 9, 2013 at 17:39
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
2
The Y-axis in a kernal density estimate of the PDF of a random variable is interpreted as a probability differential, like in calculus. It's a unitless measurement but any mesaurable area under the curve can be interpreted as a probability estimate.
-
3$\begingroup$ so given yaxis being 0.07, what does this mean in plain English. I understand histogram frequency charts, but this seems complicated to me. Can you clarify this little bit more? $\endgroup$ Commented Apr 9, 2013 at 17:18
-
$\begingroup$ The linked CV question from Huber should address your technical knowledge gap here. The literal interpretation of the Y-axis is a dimensionless quantity, like an infinitesimal grain of probability sand. An interpretation of a smoothed density estimate that's appropriate for applied audiences and scholarly journals is to calculate estimated "probability area" in pre-specified ranges. If the X-axis here were times to recurrence of cancer in months following therapy, you might be interested in risk of 1 year recurrence which you could estimate by using the function
integrate(...). $\endgroup$– AdamOCommented Apr 9, 2013 at 21:26