to use
coef(lm1)[x] rather than lm1$coefficients[x], simply because there is the coef
extractor function provided.
cheers
Mike Dunbar
>>> "Stephen Thackeray" <[EMAIL PROTECTED]> 02/05/2008 08:55:10 >>>
Dear all,
I suspect that this might be a bit basic, but
Exactly: STEER WELL CLEAR OF attach(). This is just the sort of problem you can
get into.
Mike
>>> tyler <[EMAIL PROTECTED]> 04/07/2008 00:11 >>>
"Olivia LeDee" <[EMAIL PROTECTED]> writes:
> Would someone mind helping with the following code:
>
> 1) After attaching the primary dataset
>
> ambi1
Hi Joe
I think the command you want is probably simpler than you think:
lme(HSP~coast*MBL, random= ~1|site)
or
lme(HSP~coast+MBL, random= ~1|site)
coast and MBL have distinct levels so are fixed and site is random as you say.
Having site as random will take into account that there ar
in the way that the F
ratios are calculated (i.e. what is used to construct the
denominator). Since site is random, MS coast should be compared to MS
site, rather than MS residual. Again, that's at least from my
understanding of how it works in the old-school way, perhaps the REML
fittin
(apologies - I should have written coast * MBL not ML)
I'm not sure of my ground here, but surely do lose something - you wouldn't
retain coast:MBL if it's not significant, as you lose degrees of freedom, and
this gets worse the more terms and the more interactions you consider. I think
it's a
Dear Charles
It's very difficult to tell without more information on what these papers are
and what exactly they are saying.
Also with your correlation structure, what are you trying to model? corCAR is a
continuous AR structure, allowing fractional time points, but you don't have a
time varia
lose in time are
likely correlated. I think the within group order would be proper as it
would be the same as the order of year (or tree age if you like, logyearcorr
in the model). Would the regular AR be more appropriate ?
On Tue, Nov 11, 2008 at 2:55 PM, Mike Dunbar <[EMAIL PROTECTED]> wrote:
Dear Jude
I'm trying to reconcile where you say "it doesn't meet the assumption of
homogeneity of variances" with "sample size within fields is small and
unequal". With small sample sizes you can easily generate dummy samples with
widely differing variances even where the process generating the
>From a quick look only, I'm not convinced that the structure of the model as
>suggested by Kate, is correct.
There is only one random effect in the model, site. Year and Week are specified
as random slopes. But they are not in the fixed part of the model, hence I
don't think this makes sense.