Displaying 2 results from an estimated 2 matches for "model12".
Did you mean:
model1
2010 Feb 11
1
Zero-inflated Negat. Binom. model
...I is included in blocks 1 of output containing negative binomial
regression coefficients for CAPI, and is also included also in block 2
corresponding to the inflation model. Does this make sense?
If I specify my model slightly differently, I get what I believe is more
reasonable results:
> model12 <- zeroinfl(Int_Cesto ~ 1|CAPI, dist = "negbin", EM = TRUE)
> model12
Call:
zeroinfl(formula = Int_Cesto ~ 1 | CAPI, dist = "negbin", EM = TRUE)
Count model coefficients (negbin with log link):
(Intercept)
2.692
Theta = 0.4346
Zero-inflation model coeffici...
2012 Jun 19
1
Possible bug when using encomptest
...d!
##Force encomptest to break:
encomptest(models12[[1]],models1[[1]]) ##horrible failure!
##End demonstration:
#######################################################################
Traceback and investigation:
---------------------------------------
This happens because model1 is contained in model12. This error is raised
at the
line containing the code
fm <- as.formula(paste(". ~ . +", fm))
in the encompassing test implementation in lmtest.
Suggested fix:
-------------------
This could be avoided if additional error checking is added into
encomptest(): check if
provided m...