[R-sig-ME] {dsm} warning message "step size truncated due to divergence"

[Ben Bolker]

On 14-05-25 11:06 AM, Ricardo Lima wrote:
> Hello,
> 
> I'm trying to fit a density surface model, following "Spatial models for
> distance sampling data: recent developments and future directions" (Miller
> et al. 2013 - Methods in Ecology and Evolution 4 (11): 1001-1010).
> 
> When I run the model, I get the following:
> "warning messages:
> 1: Step size truncated due to divergence
> 2: Step size truncated due to divergence
> 3: Step size truncated due to divergence
> 4: Step size truncated due to divergence
> 5: Step size truncated due to divergence "
> 
> Then, when I try to use "predict", 17 grid cells (belonging to 5 groups)
> appear as "Inf", instead of having a value. And that keeps from mapping the
> predicted abundance or calculating the abundance for the entire survey area.
> 
> Can you please help me understanding what this means and how I can get
> around this?
> 

  Pretty much impossible to say without more detail.  I could go read
the paper, but it would be more efficient if you could give a minimal
reproducible example ...  It sounds like your model is overspecified
somehow ...

  Ben Bolker