Displaying 2 results from an estimated 2 matches for "cosday".
2009 Feb 18
1
using stepAIC with negative binomial regression - error message help
...eed, temperature, barometric pressure), and tides (dummy variables for high, falling, low, and rising). So predictors are binary, categorical, and continuous variables. I ran glm.nb to fit my full model containing all first order terms. The model I ran was:
glm.nb.full<glm.nb(mantas~site+year+cosday+sinday+daylength+lunarpercent+sstmean+sststd+
sshmean+sshstd+cosdir+sindir+spd+temp+alt+tideht+high+falling+low+plankton)
However, when I use stepAIC on the model I get the message:
> stepAIC(glm.nb.full)
Start: AIC=19240.46
mantas ~ site + year + cosday + sinday + daylength + lunarpercent...
2009 Feb 19
0
using stepAIC with negative binomial regression - error message
...e "na.action" but that didn't seem to help. I didn't know of the na.omit function. It made it much easier to get rid of all the missing values in the file, and now the regression model is running.
I hadn't put interactions in the model yet since it wasn't even running. Cosday, sinday, and daylength are strongly collinear. The full model with interactions keeps all three in, plus all interactions between them.
Thanks for your help!
Tim
[[alternative HTML version deleted]]