tlyons4 wrote:I'm working on band-recovery using the Seber model. Birds are banded as HY and AHY, I have ~100-300 per each age class per year, for 15 years. When I try to fit the same age-structure to S and r that includes a difference between HY birds and others, the recovery estimate for HY birds always gets pinned at 1. I just wanted to clarify if this was a structural issue vs. some artifact of my data?
INPUT --- model={ S(~is.hy)r(~is.hy) };
INPUT --- group=1 S rows=15 cols=15 Triang ;
INPUT --- 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 1 ;
INPUT --- 1 1 1 1 1 ;
INPUT --- 1 1 1 1 ;
INPUT --- 1 1 1 ;
INPUT --- 1 1 ;
INPUT --- 1 ;
INPUT --- group=2 S rows=15 cols=15 Triang ;
INPUT --- 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 1 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 1 ;
INPUT --- 2 1 1 1 1 ;
INPUT --- 2 1 1 1 ;
INPUT --- 2 1 1 ;
INPUT --- 2 1 ;
INPUT --- 2 ;
INPUT --- group=1 r rows=15 cols=15 Triang ;
INPUT --- 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 3 ;
INPUT --- 3 3 3 3 3 ;
INPUT --- 3 3 3 3 ;
INPUT --- 3 3 3 ;
INPUT --- 3 3 ;
INPUT --- 3 ;
INPUT --- group=2 r rows=15 cols=15 Triang ;
INPUT --- 4 3 3 3 3 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 3 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 3 ;
INPUT --- 4 3 3 3 3 ;
INPUT --- 4 3 3 3 ;
INPUT --- 4 3 3 ;
INPUT --- 4 3 ;
INPUT --- 4 ;
ganghis wrote:I'm probably not thinking clearly, but I'm having trouble seeing why you wouldn't be able to estimate all 4 parameters. Clearly the adult (AHY) parameters are estimable, and then it seems like the HY would be through e.g. the number of recoveries of HY birds that survive their first year but are subsequently recovered (the expectation is R_hy * S_hy * (1-S_ahy) * r_ahy which only depends on 1 unknown; R_hy being the number of HY releases). Do you have many of these types of recoveries? It could be that the parameters are structurally identifiable but not estimable due to data sparseness.
Paul
REAL1 0.410 0.1337
REAL2 0.849 0.0150
REAL3 0.700 0.0067
REAL4 0.287 0.1476
REAL5 0.838 0.1398
REAL6 0.550 0.0082
gwhite wrote:Evan:
You're not thinking correctly. The model is just the {S(age) r(age)} model in the NPMALES example distributed with MARK. Further, this is one of the models in Program BROWNIE and described in the Brownie et al. band recovery books.
Gary
1 2 2 2 2
1 2 2 2
1 2 2
1 2
1
2 2 2 2 2
2 2 2 2
2 2 2
2 2
2
1 2 2 2 2
1 2 2 2
1 2 2
1 2
1
3 3 3 3 3
3 3 3 3
3 3 3
3 3
3
11000000000000 1194 851;
10000000000000 7589 7290;
10000001000000 196 315;
10010000000000 345 660;
10000100000000 241 430;
10000000000001 121 92;
10000000010000 171 200;
10000000000100 143 162;
00100000000000 7809 7416;
00110000000000 1161 872;
00100100000000 271 608;
00100000010000 207 316;
00100001000000 257 411;
00100000000001 136 147;
00100000000100 159 230;
00001000000000 7921 7464;
00001100000000 1149 894;
00001001000000 329 655;
00001000000100 201 311;
00001000010000 238 431;
00001000000001 162 245;
00000010000000 8023 7736;
00000011000000 1197 919;
00000010010000 318 618;
00000010000100 259 446;
00000010000001 203 281;
00000000110000 1157 887;
00000000100000 8279 8046;
00000000100001 252 385;
00000000100100 312 682;
00000000001000 8556 8488;
00000000001100 1135 913;
00000000001001 309 599;
00000000000010 8875 9141;
00000000000011 1125 859;
Real Function Parameters of {S(HY - dot/dot, AHY - dot)r(HY - dot/dot AHY - dot)}
95% Confidence Interval
Parameter Estimate Standard Error Lower Upper
-------------------------- -------------- -------------- -------------- --------------
1:S 0.6808783 18.597501 0.2968076E-072 1.0000000
2:S 0.8146861 0.0087082 0.7970079 0.8311509
3:S 0.7009760 0.0042597 0.6925608 0.7092576
4:r 0.3634081 21.178386 0.6782752E-078 1.0000000
5:r 0.2458439 6.7149868 0.4828466E-031 1.0000000
6:r 0.2973238 0.0030336 0.2914124 0.3033038
Users browsing this forum: No registered users and 10 guests