I need to remove certain interactions and keep only the one between the second
level of the factor and the continuous var t2
bin4 <- glm(resp2~ t*t2+c5.vrm,data=dfa,family="quasibinomial")
> summary(bin4)
Call:
glm(formula = resp2 ~ t * t2 + c5.vrm, family = "quasibinomial",
data = dfa)
Deviance Residuals:
Min 1Q Median 3Q Max
-6.5464 -3.0720 -1.8135 0.4896 28.9207
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) -1.492379 0.147779 -10.099 < 2e-16 ***
t2 1.381409 0.193110 7.153 1.19e-11 ***
t3 0.223979 0.275389 0.813 0.416898
t4 -1.751518 0.495391 -3.536 0.000494 ***
t5 -2.455172 0.913674 -2.687 0.007747 **
t6 -2.347207 1.131134 -2.075 0.039120 *
t2 -0.030819 0.013144 -2.345 0.019914 *
c5.vrm 0.427088 0.140085 3.049 0.002574 **
t2:t2 -0.015909 0.007292 -2.182 0.030174 *
t3:t2 -0.016170 0.010479 -1.543 0.124218
t4:t2 0.024352 0.016900 1.441 0.151009
t5:t2 -0.012071 0.034230 -0.353 0.724680
t6:t2 -0.041797 0.047106 -0.887 0.375880
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
(Dispersion parameter for quasibinomial family taken to be 41.75819)
Trying with - t3:t2 etc did not work neither did putting quotes around the
terms.
I need to have the continuous var c5.vrm same for all levels of the factor so
estimating separate equations for each level will not work. I could use offset,
but seems quite an ugly solution, there should be a way to specify exactly
which interactions are desired.
Thank you all.
Stephen Bond
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
