[Ipmpack-users] NA and lambda problems

[Eelke Jongejans]

Dear Elena,

thank you for your question and for showing a subset of your data.

ad 2: It might be that you have rethink how you are breaking down sexual 
reproduction in your system into separate steps. How are your 'number of 
flowers' and 'number of fruits' variables related? Does every fruit 
result in a new, established plant by the end of the time step? If so, 
then I can imagine that your population is exploding in size.

On the other hand, if the number of fruits per flower is a constant, it 
should be added with the 'fecConstants' argument in 'makeFecObj'. Also 
the per-seed-seedling-establishment-rate can be added in 'fecConstants'. 
Good that you studied the second chapter of the appendix to Merow et al. 
2014, which contains an example with IPMpack-code that seems relevant 
for your species as well: http://tinyurl.com/ndkh99r

Perhaps it is a good idea to discuss your species' life cycle with your 
advisors.

best wishes,
Eelke

PS: ad 1: It might be that you have rather few deaths in your dataset, 
making it hard for binomial regression models to fit all parameters. 
However, it might also be that you are having difficulties with reading 
your data into R, unrelated to IPMpack.


On 09/02/15 18:04, Elena S. wrote:
> Hi everybody,
>
> Congratulation for the pack, is very useful.
> I'm a PhD student that use IPMpack to analyze different populations of 
> a perennial plant species(so I have 3 years of data for every 
> population).
>
> I have two questions:
> 1) in one of the populations I have many NA, both in size and in 
> sizeNextcolumns, consequently also in fec0 (probability of flowering) 
> and fec1 (number of fruit).
> In the construction of models of survival and growth (via 
> survModelComp and growthModelComp) I get the following error:
> "Error in plot.window (...): need finite 'ylim' values
> In addition: There were 16 warnings (use warnings () to see them)."
> Here the code (which works well with other populations, except with this):
> survModelComp (dataf = dhsr, makePlot = TRUE, legendPos = 
> "bottomleft", Maintitle = "Survival")
> It is possible that the many NAs create a problem?
>
> 2) in all analyzed populations lambda get very high as many variables 
> added in fecundity object.
> I tried inserting the data frame columns:
> fec0 (probability of flowering), fec1 (number of fruits), nFl (number 
> of flowers).
> I get a lambdanear 1 only if I leave exclusively fec0 (the probability 
> of flowering), does anyone know how to solve by adding at least fec1?
> R0 also get excessively high values... I think that is for the same 
> problem with lambda.
>
> Thanks to all for any suggestion,
>
> Best wishes,
>
> Elena Sulis
>
>
> _______________________________________________
> IPMpack-users mailing list
> IPMpack-users at lists.r-forge.r-project.org
> http://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/ipmpack-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/ipmpack-users/attachments/20150211/1ffa59e4/attachment.html>