I have 'singular' issues for some parameters because of small populations.
Thus, I want to fix some parameters (survival in this example) to the s. value.
This code, which seems like it matches examples, adds another parameter rather than fixing the one I want. I see in the design matrix that I am not quite doing what I want.
Any help is much appreciated!
Code:
- Code: Select all
timint = c(0,1,0,1,0,0,0,0,0,0,1,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,1,0,0,0,0,1,0,1,0,0,0,0)
P.PCdotshared = list(formula=~1, share=T) #parameter constant but shared
GPGDP.fixed = list(formula=~1, fixed=0) #fix parameter to 0
s.time.fixed=list(formula=~time,fixed=list(index=c(6),value=c(0.4)))
tst1=mark(data = rob_m14col, model = "RDHuggins", time.intervals=time.intervals,
model.parameters=list(S=s.time.fixed,GammaDoublePrime=GPGDPfixed, GammaPrime=GPGDPfixed,p=PCdotshared), filename=paste('tst','3',sep=''))
- Code: Select all
From out file:
INPUT --- fixed=2;
INPUT --- parm(11)=0.4 ;
INPUT --- parm(12)=0 ;
INPUT --- group=1 S rows=9 cols=9 Triang ;
INPUT --- 1 2 3 4 5 11 7 8 9 ;
INPUT --- 2 3 4 5 6 7 8 9 ;
INPUT --- 3 4 5 6 7 8 9 ;
INPUT --- 4 5 6 7 8 9 ;
INPUT --- 5 6 7 8 9 ;
INPUT --- 6 7 8 9 ;
INPUT --- 7 8 9 ;
INPUT --- 8 9 ;
INPUT --- 9 ;
And results from out file:
- Code: Select all
LOGIT Link Function Parameters of { S(~time)Gamma''(~1)Gamma'(~1)p(~1)c() }
95% Confidence Interval
Parameter Beta Standard Error Lower Upper
------------------------- -------------- -------------- -------------- --------------
1:S:(Intercept) 0.0594393 0.5717224 -1.0611366 1.1800151
2:S:time2 0.0683505 0.7594010 -1.4200754 1.5567765
3:S:time3 -0.3472055 0.6981709 -1.7156205 1.0212096
4:S:time4 -1.2493524 0.8087020 -2.8344083 0.3357036
5:S:time5 -2.1196401 1.2068735 -4.4851123 0.2458320
6:S:time6 -10.916672 1986.0899 -3903.6530 3881.8197
7:S:time7 -34.566659 0.3890998E-007 -34.566659 -34.566659
8:S:time8 3.0709713 5.4708243 -7.6518445 13.793787
9:S:time9 -1.6043899 1.2605026 -4.0749749 0.8661952
10:p:(Intercept) 0.2142441 0.0970186 0.0240876 0.4044006
Program MARK - Survival Rate Estimation with Capture-Recapture Data
gfortran(Win64) Vers. 8.2 Sep 2017 23-Feb-2019 23:32:42 Page 006
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Real Function Parameters of { S(~time)Gamma''(~1)Gamma'(~1)p(~1)c() }
95% Confidence Interval
Parameter Estimate Standard Error Lower Upper
-------------------------- -------------- -------------- -------------- --------------
1:S g1 c1 a0 t1 0.5148554 0.1428044 0.2570923 0.7649505
2:S g1 c1 a1 t2 0.5319040 0.1123405 0.3193935 0.7334387
3:S g1 c1 a2 t3 0.4285508 0.0982207 0.2546680 0.6220683
4:S g1 c1 a3 t4 0.2332745 0.1022764 0.0902403 0.4827263
5:S g1 c1 a4 t5 0.1130257 0.1065646 0.0156172 0.5058092
6:S g1 c2 a4 t6 0.1926439E-004 0.0382601 0.1899443E-308 1.0000000
7:S g1 c1 a6 t7 0.1032060E-014 0.5900518E-015 -.1244415E-015 0.2188562E-014
8:S g1 c1 a7 t8 0.9581299 0.2183639 0.5320681E-003 0.9999990
9:S g1 c1 a8 t9 0.1758168 0.1627350 0.0230647 0.6584100
10:p g1 s1 t1 0.5533571 0.0239784 0.5060216 0.5997445
11:S g1 c1 a5 t6 0.4000000 0.0000000 0.4000000 0.4000000 Fixed
12:Gamma'' g1 c1 a0 t1 0.0000000 0.0000000 0.0000000 0.0000000 Fixed