www.phidot.org

Good day,
In a recent analysis of black bear data, a linear relationship with y-coordinates was the best-suppoted density model. Density increased as we moved North along a peninsula, away from areas of more intensive human use.

Beta parameters (coefficients)
beta SE.beta lcl ucl
D -7.4913281 0.2478314 -7.9770687 -7.0055875
D.y 0.5434952 0.1767871 0.1969988 0.8899916

Y-coordinates of mesh points ranged from 4922354 to 5011961, i.e. the region of integration extended nearly 90 km from north to south.

How can I calculate expected density at diferent locations (y-coordinates)? Does secr.fit standardize or round-off y-coordinates?

Thanks in advance,

Eric

Hi Eric
Have you met predictDsurface()? Coefficients from secr.fit relate to the normalised mask x- and y-coordinates (mean and SD stored as attribute meanSD), so manual computation is a little tricky.
Hope this is enough - let me know if you need more detail.
Murray

Code: Select all

 attr(make.mask(make.grid()), 'meanSD')


Ah.
Would you believe I'd already used PredictDsurface for plotting, but didn't think to look for the values there?
Thanks,
Eric