I had no idea this was coming up - can't believe it didn't come up in any
of my searches. Looks super useful, and I'm sure I'll be using it sooner
than later, whether on this analysis or not. The worked example is
ridiculously useful - so glad to have it.
If I'm reading the help page correctly
What other GLMM options do I have, under the restrictions of count data
with lots of zeroes and spatial autocorrelation? I was under the impression
that glmmPQL was my only choice?
On Thu, Jan 11, 2018 at 8:18 AM, Roger Bivand wrote:
> On Thu, 11 Jan 2018, Sima Usvyatsov
On Thu, 11 Jan 2018, Sima Usvyatsov wrote:
Thank you so much for your response.
Yes, I managed to muck up the fake data in 2 (!) ways - the 1,000 lons and
the fact that the lon/lats weren't repeated. Here's the correct structure.
df <- data.frame(Loc = as.factor(rep(1:20, each = 5)), Lat =
On Wed, 10 Jan 2018, Sima Usvyatsov wrote:
Hello,
I am running a negative binomial model (MASS) on count data collected on a
grid. The dataset is large - ~4,000 points, with many predictors. Being
counts, there are a lot of zeroes. All data are collected on a grid with 20
points, with high