set capture p(s) of two events=0 for some occasions

questions concerning analysis/theory using programs M-SURGE, E-SURGE and U-CARE

set capture p(s) of two events=0 for some occasions

Postby simone77 » Fri Nov 18, 2011 6:50 am

Sorry for the long title but I could not find anything more synthetic than this.

I have:
2 groups,
0 individual covariates,
3 states (including the hidden state "dead"),
4 events (0=not seen, 1=no antibodies, 2=antibodies, 3=unknown)
1 age class
9 occasions

Given I have unequal time intervals I set the age class (sensu time since first capture) =1.

At the 5th occasion no individuals were blood sampled so I have just captures of "unknown" individuals (event 3) and as a consequence, pevent1=0 and pevent2=0.

The gemaco code I have set for event step 1 (capture) is f.t (*below a question about this), i.e. capture probability depends on the interaction between individual state at the current occasion and time. I have tried to set these p=0 in the IVFV but I must been doing something wrong because the results are nonsensical (all the parameters' estimates are=0.5 that is the initial constant value set for every parameter, except those I have fixed to a different value, in the IVFV).
This is what I have done:
Image
The beta#1 is fixed to one given the first capture of each individuals has p=1 as explained in the manual (page 58). I set the beta#9 and beta#10 to zero because I understood that from 1 to 2 for T=5 and from 2 to 3 for T=5 index the parameters I am looking for, i.e. the p one individual in state 1 (no antibodies) is seen at occasion 5 and the p one individual in state 2 (antibodies) is seen at occasion 5.
But I must be missing the point....

* Should I put brackets also when this is the only sentence for a specific step? I ask this because I have read a post where Rémi Choquet suggested to put brackets around an entire sentence in order to avoid problems of priority in the sentence:
Choquet wrote:ad a(33:36).g(1)&a(22:25).g(2)&a(9:12).g(3)&a(1:4).g(4)

It should be replaced by:

ad [a(33:36).g(1)&a(22:36).g(2)&a(9:36).g(3)&a(1:36).g(4)]

because individuals of g(4) remains adults for all ages (time elapsed since first capture).
I added also bracket to avoid problems of priority in the sentence, do the same for
all shortcuts.


Thanks for any help

Simone
simone77
 
Posts: 197
Joined: Mon Aug 10, 2009 2:52 pm

Re: set capture p(s) of two events=0 for some occasions

Postby simone77 » Mon Nov 21, 2011 8:51 am

An answer to myself that might be useful for someone who, like me, is just beginning to use this software.
simone77 wrote:At the 5th occasion no individuals were blood sampled so I have just captures of "unknown" individuals (event 3) and as a consequence, pevent1=0 and pevent2=0.
.
That's right, but I was setting pevent1=0 and pevent2=0 in the wrong pattern matrix (event step one -> capture matrix) while I had to set those null probabilities in the pattern matrix event step two (state assignment).
In fact the probability an individual in state 1 or 2 was captured in that session is not null but it is null the probability an individual in state 1 or 2 was seen in state 1 or 2 (state assignment).
simone77
 
Posts: 197
Joined: Mon Aug 10, 2009 2:52 pm


Return to analysis help

Who is online

Users browsing this forum: Google [Bot] and 16 guests