TMSURVIV
User's Manual
by
James E. Hines
Biological Resources Division, USGS(1)
11510 American Holly Dr. #201
Patuxent Wildlife Research Center
Laurel, MD 20708-4017
email: jim_hines%40usgs.gov
Introduction
Program TMSURVIV (Transient-Model-SURVIVal analysis) computes parameter estimates
of survival and capture probability and proportions of "residents" under the Jolly-Seber models
described in "Capture-Recapture Survival Models Taking Account of Transients" (Pradel et. al.,
1997). Actually, TMSURVIV is a specially modified version of Dr. G. White's program
SURVIV (White, 1983) which incorporates the transient models. With this program and it's
companion program, CNVTMSRV, users are able to get parameter estimates for these complex
models from capture-history data without having to specify the cell probabilities.
TMSURVIV is intended to be used in a situation where the sampling area includes animals
which pass through the area and are not part of the "resident" population. .
Output from TMSURVIV includes survival probability estimates, capture probability estimates, estimates of the proportion of residents, goodness-of-fit tests, and likelihood-ratio tests.
By default, models are generated for standard Jolly-Seber models (where all animals are
assumed to be residents), and transient models (a portion of the population is transient and only
available for capture in one sample). The model names signify the structure of the model. Each
model name is in the form "S.P.G." where "S" stands for survival, "P" stands for capture
probability, and "G" stands for "gamma (proportion of residents)". If S,P, or G are followed by
a "t", then the corresponding parameter is "time-dependent". If there is no "G" in the name, then
all animals are assumed to be residents (proportion of residents = 1.0), which is equivalent to the
Jolly-Seber models.
Since program TMSURVIV is a generalization of program JOLLY, the conversion/model generation program (CNVTMSRV) will read input files designed for program JOLLY. Input consists of capture-histories for each animal, preceded by statements defining the specifics of the data (title, number of periods, interval lengths).
Using CNVTMSRV
To run CNVTMSRV, type CNVTMSRV at the DOS prompt and respond to the program
prompts. When the program is run, you are prompted for the name of the file(s) containing the
capture-histories. Analysis on groups of animals may be performed by separating the capture-histories into different files. CNVTMSRV will prompt for each input file and a 1-character
identifier (eg. M for male, F for female) until 'END' is entered for the input file.
Here is the output file created by CNVTMSRV which can be input to TMSURVIV:
PROC TITLE 'Lazuli bunting data from DESIGN & ANAL. METH... Burhnam et al.'; PROC MODEL NPAR=21 ADDCELL; COHORT= 168 /* unmarked in 1 */; 31:; 4:; 0:; 1:; 0:; 0:; 0:; COHORT= 31 /* marked in 2 */; 7:; 4:; 0:; 0:; 0:; 0:; COHORT= 367 /* unmarked in 2 */; 12:; 10:; 4:; 1:; 0:; 1:; COHORT= 23 /* marked in 3 */; 12:; 0:; 1:; 0:; 0:; COHORT= 65 /* unmarked in 3 */; 8:; 3:; 0:; 0:; 0:; COHORT= 34 /* marked in 4 */; 16:; 3:; 0:; 0:; COHORT= 230 /* unmarked in 4 */; 25:; 3:; 3:; 1:; COHORT= 49 /* marked in 5 */; 20:; 0:; 0:; COHORT= 255 /* unmarked in 5 */; 38:; 13:; 1:; COHORT= 66 /* marked in 6 */; 28:; 2:; COHORT= 256 /* unmarked in 6 */; 39:; 7:; COHORT= 83 /* marked in 7 */; 30:; COHORT= 240 /* unmarked in 7 */; 46:; LABELS; S(1)=PHI(1); S(2)=PHI(2); S(3)=PHI(3); S(4)=PHI(4); S(5)=PHI(5); S(6)=PHI(6); S(7)=PHI(7); S(8)=p(2); S(9)=p(3); S(10)=p(4); S(11)=p(5); S(12)=p(6); S(13)=p(7); S(14)=p(8); S(15)=gamma(1); S(16)=gamma(2); S(17)=gamma(3); S(18)=gamma(4); S(19)=gamma(5); S(20)=gamma(6); S(21)=gamma(7); PROC ESTIMATE NOVAR MAXFN=32000 NAME=SP; CONSTRAINTS; S(2)=S(1); S(3)=S(1); S(4)=S(1); S(5)=S(1); S(6)=S(1); S(7)=S(1); S(9)=S(8); S(10)=S(8); S(11)=S(8); S(12)=S(8); S(13)=S(8); S(14)=S(8); S(15)=1; S(16)=1; S(17)=1; S(18)=1; S(19)=1; S(20)=1; S(21)=1; PROC ESTIMATE NOVAR MAXFN=32000 NAME=SPG; CONSTRAINTS; S(2)=S(1); S(3)=S(1); S(4)=S(1); S(5)=S(1); S(6)=S(1); S(7)=S(1); S(9)=S(8); S(10)=S(8); S(11)=S(8); S(12)=S(8); S(13)=S(8); S(14)=S(8); S(16)=S(15); S(17)=S(15); S(18)=S(15); S(19)=S(15); S(20)=S(15); S(21)=S(15); PROC ESTIMATE NOVAR MAXFN=32000 NAME=StPt; CONSTRAINTS; S(16)=1; S(17)=1; S(18)=1; S(19)=1; S(20)=1; S(21)=1; PROC ESTIMATE NOVAR MAXFN=1 NAME=StPtGt; CONSTRAINTS; S(14)=1; S(15)=1; initial; s(1)=0.588756 ; s(2)=0.480049 ; s(3)=0.727235 ; s(4)=0.601595 ; s(5)=0.608575 ; s(6)=0.501442 ; s(7)=0.349755 ; s(8)=0.313413 ; s(9)=0.454870 ; s(10)=0.655431 ; s(11)=0.688543 ; s(12)=0.629337 ; s(13)=0.737137 ; s(16)=0.168890 ; s(17)=0.214131 ; s(18)=0.286596 ; s(19)=0.364735 ; s(20)=0.453364 ; s(21)=0.562283 ; PROC TEST; PROC STOP;
The output file, CNVTMSRV.OUT, contains a title statement, statements to define the input data, label definitions, statements to describe models, a test statement and a stop statement.
The title statement is used to identify the data used in the analysis.
The model definition statements start with "PROC MODEL NPAR=21 ADDCELL;" and
end with "COHORT=240;". The "NPAR=21" indicates the maximum number of parameters
(estimated or fixed) in any of the following models. In this example there are 7 survival rate
parameters, 7 capture probability parameters, and 7 residency parameters.
Following the data are the labels for each of the parameters. Internally, the parameters
are called "S(1), S(2), ... S(NPAR). The labels relate these internal parameters to meaningful
labels for these models. The survival parameters are labelled PHI, followed by time period in
parenthesis. The capture-probability parameters are labelled "p", and the residency parameters
are labelled "gamma". If more than one group was input, the label will have a 1-character
identifier tacked onto the end to indicate the group.
After the label definitions come the model definitions for each model. Each model starts
with a "PROC ESTIMATE" statement. Options on the "PROC ESTIMATE" statement include
"NOVAR" which inhibits printing of the variance-covariance matrix, the number of significant
digits, "NSIG" (i.e., number of digits following the decimal point which do not change at the end
of the iterative process), the maximum number of function evaluations, "MAXFN", and the name
of the model. If the variance-covariance matrix of parameter estimates is desired, delete the string
"NOVAR" using a text editor.
CNVTMSRV produces the model definitions from most restrictive (model "SP") to most
general (model "StPtGt"). The reason for this is that the most restrictive model has the fewest
estimable parameters and converges more easily. Final estimates from this model can then be
input as starting values to more general models. If more than one group was entered in
CNVTMSRV, a series of additional models is produced which assumes the data came from only
one group (eg. 1SP, 1StPtGt,...).
The statements following the "CONSTRAINTS" statement describe each model in terms
of the most general model. In model "SP", the survival and capture probabilities are assumed
to be constant across time, so the survival probabilities for time period two and three are set equal
to the survival probabilities for time period one. The capture probabilities are also assumed
constant over time, so the capture probabilities for time periods 3 and 4 are set equal to the
capture probabilities for time period 2. These equalities must be specified in terms of parameter
number which can be obtained from the labels section.
The sequence of statements starting with "PROC ESTIMATE ... NAME=SPG" cause
TMSURVIV to produce estimates under the model with constant survival and capture probabilities
and a constant proportion of residents.
The sequence of statements starting with "PROC ESTIMATE ... NAME=StP" cause
TMSURVIV to produce estimates under the model with time-specific survival probabilities and
constant capture probabilities (equivalent to model "B" in JOLLY).
The "PROC TEST;" statement causes TMSURVIV to print tables of statistics used for comparison of the models. "PROC STOP;" causes TMSURVIV to stop execution even if more statements follow.
TMSURVIV
TMSURVIV prompts for one line of input to specify the name of the input and output files
and command line options. When the program is run, the following prompt appears:
Enter command line parameters [i=in_file] [l=out_file]
[lines=n] [compile run] [noecho]:
At this prompt, any or all of the items enclosed in brackets may be specified. If
"i=in_file" is specified, the input will be read from the file "in_file". Usually, this is the file
created by CNVTMSRV and is called CNVTMSRV.OUT unless it has been renamed. A full
pathname may be used to indicate a different directory. If this item is omitted, TMSURVIV
expects the input from the keyboard. (Cntl-Break will abort the program).
If "l=out_file" is specified, output from TMSURVIV will be directed to the file
"out_file". The default output file is the CRT screen. To direct output directly to the printer, use
"l=lpt1".
If "lines=n" is included, TMSURVIV will print a header and the title in the output file
after every n lines. The default value for n is 60.
If "compile run" is included, TMSURVIV will only check the input file for errors and not
perform any analysis. This option is used in SURVIV to create the estimation routine and is not
needed in TMSURVIV.
The "noecho" option causes TMSURVIV to suppress printing of the input data. This
option is useful when there are several runs of models on the same data and you would like to
conserve paper, but at least one run should contain a listing of the data to check for "typos".
To run the sample data file with TMSURVIV, enter the following at the above prompt:
i=cnvtmsrv.out l=sample.out lines=9999
The output produced by TMSURVIV contains a listing of the input data, estimates of the parameters under each model, a goodness-of-fit test for each model, an AIC statistic for each model, and between model tests. The following output was created using TMSURVIV on the sample data file listed previously:
SURVIV - Survival Rate Estimation with User Specified Cell Probabilities
2-Sep-97 15:47:55 Version 1.4(OS/2) June, 1991 Page 001
INPUT --- PROC TITLE 'Lazuli bunting data from DESIGN & ANAL. METH...
INPUT --- Burhnam et al.';
CPU time in seconds for last procedure was 0.00
INPUT --- PROC MODEL NPAR=21 ADDCELL;
INPUT --- COHORT= 168 /* unmarked in 1 */;
INPUT --- 31:;4:;0:;1:;0:;0:;0:;
INPUT --- COHORT= 31 /* marked in 2 */;
INPUT --- 7:;4:;0:;0:;0:;0:;
INPUT --- COHORT= 367 /* unmarked in 2 */;
INPUT --- 12:;10:;4:;1:;0:;1:;
INPUT --- COHORT= 23 /* marked in 3 */;
INPUT --- 12:;0:;1:;0:;0:;
INPUT --- COHORT= 65 /* unmarked in 3 */;
INPUT --- 8:;3:;0:;0:;0:;
INPUT --- COHORT= 34 /* marked in 4 */;
INPUT --- 16:;3:;0:;0:;
INPUT --- COHORT= 230 /* unmarked in 4 */;
INPUT --- 25:;3:;3:;1:;
INPUT --- COHORT= 49 /* marked in 5 */;
INPUT --- 20:;0:;0:;
INPUT --- COHORT= 255 /* unmarked in 5 */;
INPUT --- 38:;13:;1:;
INPUT --- COHORT= 66 /* marked in 6 */;
INPUT --- 28:;2:;
INPUT --- COHORT= 256 /* unmarked in 6 */;
INPUT --- 39:;7:;
INPUT --- COHORT= 83 /* marked in 7 */;
INPUT --- 30:;
INPUT --- COHORT= 240 /* unmarked in 7 */;
INPUT --- 46:;
INPUT --- PROC ESTIMATE NOVAR MAXFN=32000 NAME=SP;
Number of parameters in model = 21
Number of parameters set equal = 12
Number of parameters fixed = 7
Number of parameters estimated = 2
Final function value 1133.1718 (Error Return = 0)
Number of significant digits 7
Number of function evaluations 45
95% Confidence Interval
I Parameter S(I) Standard Error Lower Upper
--- -------------------- ------------ ------------ ------------ ------------
1 1 PHI(1) 0.379861 0.200672E-01 0.340529 0.419193
2 1 PHI(2) 0.379861 0.200672E-01 0.340529 0.419193
3 1 PHI(3) 0.379861 0.200672E-01 0.340529 0.419193
4 1 PHI(4) 0.379861 0.200672E-01 0.340529 0.419193
5 1 PHI(5) 0.379861 0.200672E-01 0.340529 0.419193
6 1 PHI(6) 0.379861 0.200672E-01 0.340529 0.419193
7 1 PHI(7) 0.379861 0.200672E-01 0.340529 0.419193
8 2 p(2) 0.432594 0.327044E-01 0.368494 0.496695
9 2 p(3) 0.432594 0.327044E-01 0.368494 0.496695
10 2 p(4) 0.432594 0.327044E-01 0.368494 0.496695
11 2 p(5) 0.432594 0.327044E-01 0.368494 0.496695
12 2 p(6) 0.432594 0.327044E-01 0.368494 0.496695
13 2 p(7) 0.432594 0.327044E-01 0.368494 0.496695
14 2 p(8) 0.432594 0.327044E-01 0.368494 0.496695
15 -15 gamma(1) 1.00000 0.000000E+00 1.00000 1.00000
16 -16 gamma(2) 1.00000 0.000000E+00 1.00000 1.00000
17 -17 gamma(3) 1.00000 0.000000E+00 1.00000 1.00000
18 -18 gamma(4) 1.00000 0.000000E+00 1.00000 1.00000
19 -19 gamma(5) 1.00000 0.000000E+00 1.00000 1.00000
20 -20 gamma(6) 1.00000 0.000000E+00 1.00000 1.00000
21 -21 gamma(7) 1.00000 0.000000E+00 1.00000 1.00000
Cohort Cell Observed Expected Chi-square Note
------ ---- -------- -------- ---------- -------------
1 1 31 27.607 0.417 0 < P < 1
1 2 4 5.950 0.639 0 < P < 1
1 3 0 1.282 1.282 0 < P < 1
1 4 1 0.276 1.894 0 < P < 1
1 5 0 0.060 0.060 0 < P < 1
1 6 0 0.013 0.013 0 < P < 1
1 7 0 0.003 0.003 0 < P < 1
1 8 132 132.809 0.005 0 < P < 1
1 Cohort df= 2 1.303 P = 0.5214
2 1 7 5.094 0.713 0 < P < 1
2 2 4 1.098 7.670 0 < P < 1
2 3 0 0.237 0.237 0 < P < 1
2 4 0 0.051 0.051 0 < P < 1
2 5 0 0.011 0.011 0 < P < 1
2 6 0 0.002 0.002 0 < P < 1
2 7 20 24.507 0.829 0 < P < 1
2 Cohort df= 1 3.957 P = 0.0467
3 1 12 60.308 38.695 0 < P < 1
3 2 10 12.998 0.692 0 < P < 1
3 3 4 2.802 0.513 0 < P < 1
3 4 1 0.604 0.260 0 < P < 1
3 5 0 0.130 0.130 0 < P < 1
3 6 1 0.028 33.676 0 < P < 1
3 7 339 290.130 8.232 0 < P < 1
3 Cohort df= 3 49.284 P = 0.0000
4 1 12 3.779 17.880 0 < P < 1
4 2 0 0.815 0.815 0 < P < 1
4 3 1 0.176 3.871 0 < P < 1
4 4 0 0.038 0.038 0 < P < 1
4 5 0 0.008 0.008 0 < P < 1
4 6 10 18.184 3.684 0 < P < 1
4 Cohort df= 1 17.593 P = 0.0000
5 1 8 10.681 0.673 0 < P < 1
5 2 3 2.302 0.212 0 < P < 1
5 3 0 0.496 0.496 0 < P < 1
5 4 0 0.107 0.107 0 < P < 1
5 5 0 0.023 0.023 0 < P < 1
5 6 54 51.390 0.133 0 < P < 1
5 Cohort df= 2 0.807 P = 0.6679
6 1 16 5.587 19.407 0 < P < 1
6 2 3 1.204 2.678 0 < P < 1
6 3 0 0.260 0.260 0 < P < 1
6 4 0 0.056 0.056 0 < P < 1
6 5 15 26.893 5.260 0 < P < 1
6 Cohort df= 1 25.163 P = 0.0000
7 1 25 37.795 4.332 0 < P < 1
7 2 3 8.146 3.251 0 < P < 1
7 3 3 1.756 0.882 0 < P < 1
7 4 1 0.378 1.021 0 < P < 1
7 5 198 181.925 1.420 0 < P < 1
7 Cohort df= 3 10.634 P = 0.0139
8 1 20 8.052 17.729 0 < P < 1
8 2 0 1.735 1.735 0 < P < 1
8 3 0 0.374 0.374 0 < P < 1
8 4 29 38.838 2.492 0 < P < 1
8 Cohort df= 2 22.331 P = 0.0000
9 1 38 41.903 0.364 0 < P < 1
9 2 13 9.032 1.744 0 < P < 1
9 3 1 1.947 0.460 0 < P < 1
9 4 203 202.119 0.004 0 < P < 1
9 Cohort df= 2 1.199 P = 0.5490
10 1 28 10.846 27.134 0 < P < 1
10 2 2 2.338 0.049 0 < P < 1
10 3 36 52.817 5.355 0 < P < 1
10 Cohort df= 2 32.537 P = 0.0000
11 1 39 42.067 0.224 0 < P < 1
11 2 7 9.067 0.471 0 < P < 1
11 3 210 204.866 0.129 0 < P < 1
11 Cohort df= 2 0.824 P = 0.6625
12 1 30 13.639 19.626 0 < P < 1
12 2 53 69.361 3.859 0 < P < 1
12 Cohort df= 1 23.485 P = 0.0000
13 1 46 39.438 1.092 0 < P < 1
13 2 194 200.562 0.215 0 < P < 1
13 Cohort df= 1 1.306 P = 0.2530
------------------------------------------------------------
@@ 1 0 0 47 207.123 21 190.424 -157.121 318.242
G Total (Degrees of freedom = 47) 207.123
Pr(Larger Chi-square) = 0.0000
With pooling, Degrees of freedom = 21 Pearson Chi-square = 190.424
Pr(Larger Chi-square) = 0.0000
Log-likelihood = -157.12123 Akaike Information Criterion = 318.24245
CPU time in seconds for last procedure was 0.03
INPUT --- PROC ESTIMATE NOVAR MAXFN=32000 NAME=SPG;
Number of parameters in model = 21
Number of parameters set equal = 18
Number of parameters fixed = 0
Number of parameters estimated = 3
Final function value 1081.9256 (Error Return = 0)
Number of significant digits 8
Number of function evaluations 71
95% Confidence Interval
I Parameter S(I) Standard Error Lower Upper
--- -------------------- ------------ ------------ ------------ ------------
1 1 PHI(1) 0.586343 0.291084E-01 0.529291 0.643396
2 1 PHI(2) 0.586343 0.291084E-01 0.529291 0.643396
3 1 PHI(3) 0.586343 0.291084E-01 0.529291 0.643396
4 1 PHI(4) 0.586343 0.291084E-01 0.529291 0.643396
5 1 PHI(5) 0.586343 0.291084E-01 0.529291 0.643396
6 1 PHI(6) 0.586343 0.291084E-01 0.529291 0.643396
7 1 PHI(7) 0.586343 0.291084E-01 0.529291 0.643396
8 2 p(2) 0.617879 0.328420E-01 0.553509 0.682250
9 2 p(3) 0.617879 0.328420E-01 0.553509 0.682250
10 2 p(4) 0.617879 0.328420E-01 0.553509 0.682250
11 2 p(5) 0.617879 0.328420E-01 0.553509 0.682250
12 2 p(6) 0.617879 0.328420E-01 0.553509 0.682250
13 2 p(7) 0.617879 0.328420E-01 0.553509 0.682250
14 2 p(8) 0.617879 0.328420E-01 0.553509 0.682250
15 3 gamma(1) 0.354799 0.316262E-01 0.292812 0.416786
16 3 gamma(2) 0.354799 0.316262E-01 0.292812 0.416786
17 3 gamma(3) 0.354799 0.316262E-01 0.292812 0.416786
18 3 gamma(4) 0.354799 0.316262E-01 0.292812 0.416786
19 3 gamma(5) 0.354799 0.316262E-01 0.292812 0.416786
20 3 gamma(6) 0.354799 0.316262E-01 0.292812 0.416786
21 3 gamma(7) 0.354799 0.316262E-01 0.292812 0.416786
Cohort Cell Observed Expected Chi-square Note
------ ---- -------- -------- ---------- -------------
1 1 31 21.595 4.096 0 < P < 1
1 2 4 4.838 0.145 0 < P < 1
1 3 0 1.084 1.084 0 < P < 1
1 4 1 0.243 2.360 0 < P < 1
1 5 0 0.054 0.054 0 < P < 1
1 6 0 0.012 0.012 0 < P < 1
1 7 0 0.003 0.003 0 < P < 1
1 8 132 140.171 0.476 0 < P < 1
1 Cohort df= 2 4.817 P = 0.0899
2 1 7 11.231 1.594 0 < P < 1
2 2 4 2.516 0.875 0 < P < 1
2 3 0 0.564 0.564 0 < P < 1
2 4 0 0.126 0.126 0 < P < 1
2 5 0 0.028 0.028 0 < P < 1
2 6 0 0.006 0.006 0 < P < 1
2 7 20 16.528 0.729 0 < P < 1
2 Cohort df= 2 2.501 P = 0.2864
3 1 12 47.174 26.227 0 < P < 1
3 2 10 10.570 0.031 0 < P < 1
3 3 4 2.368 1.124 0 < P < 1
3 4 1 0.531 0.415 0 < P < 1
3 5 0 0.119 0.119 0 < P < 1
3 6 1 0.027 35.570 0 < P < 1
3 7 339 306.212 3.511 0 < P < 1
3 Cohort df= 3 32.638 P = 0.0000
4 1 12 8.333 1.614 0 < P < 1
4 2 0 1.867 1.867 0 < P < 1
4 3 1 0.418 0.809 0 < P < 1
4 4 0 0.094 0.094 0 < P < 1
4 5 0 0.021 0.021 0 < P < 1
4 6 10 12.267 0.419 0 < P < 1
4 Cohort df= 2 2.850 P = 0.2405
5 1 8 8.355 0.015 0 < P < 1
5 2 3 1.872 0.680 0 < P < 1
5 3 0 0.419 0.419 0 < P < 1
5 4 0 0.094 0.094 0 < P < 1
5 5 0 0.021 0.021 0 < P < 1
5 6 54 54.238 0.001 0 < P < 1
5 Cohort df= 2 0.163 P = 0.9219
6 1 16 12.318 1.101 0 < P < 1
6 2 3 2.760 0.021 0 < P < 1
6 3 0 0.618 0.618 0 < P < 1
6 4 0 0.139 0.139 0 < P < 1
6 5 15 18.165 0.552 0 < P < 1
6 Cohort df= 2 1.728 P = 0.4214
7 1 25 29.564 0.705 0 < P < 1
7 2 3 6.624 1.983 0 < P < 1
7 3 3 1.484 1.548 0 < P < 1
7 4 1 0.333 1.340 0 < P < 1
7 5 198 191.995 0.188 0 < P < 1
7 Cohort df= 2 1.138 P = 0.5660
8 1 20 17.752 0.285 0 < P < 1
8 2 0 3.977 3.977 0 < P < 1
8 3 0 0.891 0.891 0 < P < 1
8 4 29 26.379 0.260 0 < P < 1
8 Cohort df= 2 5.414 P = 0.0668
9 1 38 32.778 0.832 0 < P < 1
9 2 13 7.344 4.356 0 < P < 1
9 3 1 1.645 0.253 0 < P < 1
9 4 203 213.233 0.491 0 < P < 1
9 Cohort df= 2 4.116 P = 0.1277
10 1 28 23.911 0.699 0 < P < 1
10 2 2 5.357 2.104 0 < P < 1
10 3 36 36.732 0.015 0 < P < 1
10 Cohort df= 2 2.818 P = 0.2444
11 1 39 32.906 1.128 0 < P < 1
11 2 7 7.373 0.019 0 < P < 1
11 3 210 215.721 0.152 0 < P < 1
11 Cohort df= 2 1.299 P = 0.5223
12 1 30 30.070 0.000 0 < P < 1
12 2 53 52.930 0.000 0 < P < 1
12 Cohort df= 1 0.000 P = 0.9872
13 1 46 30.850 7.440 0 < P < 1
13 2 194 209.150 1.097 0 < P < 1
13 Cohort df= 1 8.538 P = 0.0035
------------------------------------------------------------
@@ 2 0 0 46 104.630 22 68.0196 -105.875 217.750
G Total (Degrees of freedom = 46) 104.630
Pr(Larger Chi-square) = 0.0000
With pooling, Degrees of freedom = 22 Pearson Chi-square = 68.020
Pr(Larger Chi-square) = 0.0000
Log-likelihood = -105.87508 Akaike Information Criterion = 217.75016
CPU time in seconds for last procedure was 0.05
INPUT --- PROC ESTIMATE NOVAR MAXFN=32000 NAME=StPt;
Number of parameters in model = 21
Number of parameters set equal = 0
Number of parameters fixed = 8
Number of parameters estimated = 13
Final function value 1097.6643 (Error Return = 0)
Number of significant digits 6
Number of function evaluations 466
95% Confidence Interval
I Parameter S(I) Standard Error Lower Upper
--- -------------------- ------------ ------------ ------------ ------------
1 1 PHI(1) 0.488248 0.142347 0.209248 0.767248
2 2 PHI(2) 0.222704 0.492480E-01 0.126178 0.319231
3 3 PHI(3) 0.551159 0.128118 0.300047 0.802270
4 4 PHI(4) 0.323701 0.548025E-01 0.216288 0.431114
5 5 PHI(5) 0.408204 0.589237E-01 0.292714 0.523695
6 6 PHI(6) 0.332919 0.442633E-01 0.246163 0.419675
7 7 PHI(7) 0.235294 0.236021E-01 0.189034 0.281554
8 8 p(2) 0.377931 0.118809 0.145065 0.610797
9 9 p(3) 0.230000 0.617062E-01 0.109056 0.350944
10 10 p(4) 0.373868 0.863731E-01 0.204577 0.543159
11 11 p(5) 0.471658 0.818681E-01 0.311196 0.632119
12 12 p(6) 0.450512 0.690642E-01 0.315146 0.585878
13 13 p(7) 0.619403 0.765506E-01 0.469364 0.769442
14 -14 p(8) 1.00000 0.000000E+00 1.00000 1.00000
15 -15 gamma(1) 1.00000 0.000000E+00 1.00000 1.00000
16 -16 gamma(2) 1.00000 0.000000E+00 1.00000 1.00000
17 -17 gamma(3) 1.00000 0.000000E+00 1.00000 1.00000
18 -18 gamma(4) 1.00000 0.000000E+00 1.00000 1.00000
19 -19 gamma(5) 1.00000 0.000000E+00 1.00000 1.00000
20 -20 gamma(6) 1.00000 0.000000E+00 1.00000 1.00000
21 -21 gamma(7) 1.00000 0.000000E+00 1.00000 1.00000
Cohort Cell Observed Expected Chi-square Note
------ ---- -------- -------- ---------- -------------
1 1 31 31.000 0.000 0 < P < 1
1 2 4 2.614 0.735 0 < P < 1
1 3 0 1.803 1.803 0 < P < 1
1 4 1 0.461 0.630 0 < P < 1
1 5 0 0.095 0.095 0 < P < 1
1 6 0 0.024 0.024 0 < P < 1
1 7 0 0.003 0.003 0 < P < 1
1 8 132 132.000 0.000 0 < P < 1
1 Cohort df= 3 1.541 P = 0.6729
2 1 7 1.588 18.447 0 < P < 1
2 2 4 1.095 7.702 0 < P < 1
2 3 0 0.280 0.280 0 < P < 1
2 4 0 0.058 0.058 0 < P < 1
2 5 0 0.015 0.015 0 < P < 1
2 6 0 0.002 0.002 0 < P < 1
2 7 20 27.962 2.267 0 < P < 1
2 Cohort df= 1 23.138 P = 0.0000
3 1 12 18.798 2.459 0 < P < 1
3 2 10 12.968 0.679 0 < P < 1
3 3 4 3.316 0.141 0 < P < 1
3 4 1 0.683 0.147 0 < P < 1
3 5 0 0.172 0.172 0 < P < 1
3 6 1 0.025 38.284 0 < P < 1
3 7 339 331.038 0.192 0 < P < 1
3 Cohort df= 3 4.106 P = 0.2503
4 1 12 4.739 11.123 0 < P < 1
4 2 0 1.212 1.212 0 < P < 1
4 3 1 0.250 2.255 0 < P < 1
4 4 0 0.063 0.063 0 < P < 1
4 5 0 0.009 0.009 0 < P < 1
4 6 10 16.727 2.706 0 < P < 1
4 Cohort df= 1 9.920 P = 0.0016
5 1 8 13.394 2.172 0 < P < 1
5 2 3 3.425 0.053 0 < P < 1
5 3 0 0.706 0.706 0 < P < 1
5 4 0 0.177 0.177 0 < P < 1
5 5 0 0.026 0.026 0 < P < 1
5 6 54 47.273 0.957 0 < P < 1
5 Cohort df= 2 3.540 P = 0.1703
6 1 16 5.191 22.507 0 < P < 1
6 2 3 1.069 3.486 0 < P < 1
6 3 0 0.269 0.269 0 < P < 1
6 4 0 0.039 0.039 0 < P < 1
6 5 15 27.432 5.634 0 < P < 1
6 Cohort df= 1 29.164 P = 0.0000
7 1 25 35.115 2.914 0 < P < 1
7 2 3 7.234 2.478 0 < P < 1
7 3 3 1.819 0.766 0 < P < 1
7 4 1 0.263 2.065 0 < P < 1
7 5 198 185.568 0.833 0 < P < 1
7 Cohort df= 3 7.990 P = 0.0462
8 1 20 9.011 13.401 0 < P < 1
8 2 0 2.266 2.266 0 < P < 1
8 3 0 0.328 0.328 0 < P < 1
8 4 29 37.395 1.885 0 < P < 1
8 Cohort df= 2 17.879 P = 0.0001
9 1 38 46.895 1.687 0 < P < 1
9 2 13 11.795 0.123 0 < P < 1
9 3 1 1.705 0.292 0 < P < 1
9 4 203 194.605 0.362 0 < P < 1
9 Cohort df= 2 2.068 P = 0.3556
10 1 28 13.610 15.215 0 < P < 1
10 2 2 1.968 0.001 0 < P < 1
10 3 36 50.422 4.125 0 < P < 1
10 Cohort df= 1 17.478 P = 0.0000
11 1 39 52.790 3.602 0 < P < 1
11 2 7 7.632 0.052 0 < P < 1
11 3 210 195.578 1.064 0 < P < 1
11 Cohort df= 2 4.718 P = 0.0945
12 1 30 19.529 5.614 0 < P < 1
12 2 53 63.471 1.727 0 < P < 1
12 Cohort df= 1 7.341 P = 0.0067
13 1 46 56.471 1.941 0 < P < 1
13 2 194 183.529 0.597 0 < P < 1
13 Cohort df= 1 2.539 P = 0.1111
------------------------------------------------------------
@@ 3 0 0 36 136.108 10 131.422 -121.614 269.228
G Total (Degrees of freedom = 36) 136.108
Pr(Larger Chi-square) = 0.0000
With pooling, Degrees of freedom = 10 Pearson Chi-square = 131.422
Pr(Larger Chi-square) = 0.0000
Log-likelihood = -121.61377 Akaike Information Criterion = 269.22754
CPU time in seconds for last procedure was 0.14
INPUT --- PROC ESTIMATE NOVAR MAXFN=32000 NAME=StPtGt;
Number of parameters in model = 21
Number of parameters set equal = 0
Number of parameters fixed = 2
Number of parameters estimated = 19
Final function value 1050.5491 (Error Return = 0)
Number of significant digits 6
Number of function evaluations 876
95% Confidence Interval
I Parameter S(I) Standard Error Lower Upper
--- -------------------- ------------ ------------ ------------ ------------
1 1 PHI(1) 0.268398 0.498914E-01 0.170611 0.366185
2 2 PHI(2) 0.485112 0.131254 0.227854 0.742369
3 3 PHI(3) 0.674294 0.132072 0.415433 0.933156
4 4 PHI(4) 0.728724 0.128546 0.476773 0.980675
5 5 PHI(5) 0.517647 0.956648E-01 0.330144 0.705150
6 6 PHI(6) 0.555981 0.831053E-01 0.393094 0.718867
7 7 PHI(7) 0.361446 0.527329E-01 0.258089 0.464802
8 8 p(2) 0.687500 0.115878 0.460379 0.914621
9 9 p(3) 0.382353 0.833418E-01 0.219003 0.545703
10 10 p(4) 0.633333 0.879815E-01 0.460890 0.805777
11 11 p(5) 0.606061 0.850581E-01 0.439347 0.772775
12 12 p(6) 0.612245 0.696055E-01 0.475818 0.748672
13 13 p(7) 0.714286 0.697071E-01 0.577660 0.850912
14 -20 p(8) 1.00000 0.000000E+00 1.00000 1.00000
15 -21 gamma(1) 1.00000 0.000000E+00 1.00000 1.00000
16 16 gamma(2) 0.215011 0.650886E-01 0.874375E-01 0.342585
17 17 gamma(3) 0.299408 0.988362E-01 0.105689 0.493127
18 18 gamma(4) 0.248970 0.557393E-01 0.139721 0.358219
19 19 gamma(5) 0.499608 0.105866 0.292110 0.707106
20 14 gamma(6) 0.395313 0.750206E-01 0.248272 0.542353
21 15 gamma(7) 0.530278 0.104530 0.325399 0.735157
Cohort Cell Observed Expected Chi-square Note
------ ---- -------- -------- ---------- -------------
1 1 31 31.000 0.000 0 < P < 1
1 2 4 2.614 0.735 0 < P < 1
1 3 0 1.803 1.803 0 < P < 1
1 4 1 0.461 0.630 0 < P < 1
1 5 0 0.095 0.095 0 < P < 1
1 6 0 0.024 0.024 0 < P < 1
1 7 0 0.003 0.003 0 < P < 1
1 8 132 132.000 0.000 0 < P < 1
1 Cohort df= 3 1.541 P = 0.6729
2 1 7 5.750 0.272 0 < P < 1
2 2 4 3.967 0.000 0 < P < 1
2 3 0 1.014 1.014 0 < P < 1
2 4 0 0.209 0.209 0 < P < 1
2 5 0 0.053 0.053 0 < P < 1
2 6 0 0.008 0.008 0 < P < 1
2 7 20 20.000 0.000 0 < P < 1
2 Cohort df= 2 0.569 P = 0.7523
3 1 12 14.636 0.475 0 < P < 1
3 2 10 10.097 0.001 0 < P < 1
3 3 4 2.582 0.779 0 < P < 1
3 4 1 0.532 0.412 0 < P < 1
3 5 0 0.134 0.134 0 < P < 1
3 6 1 0.019 49.727 0 < P < 1
3 7 339 339.000 0.000 0 < P < 1
3 Cohort df= 3 2.763 P = 0.4296
4 1 12 9.822 0.483 0 < P < 1
4 2 0 2.511 2.511 0 < P < 1
4 3 1 0.517 0.450 0 < P < 1
4 4 0 0.130 0.130 0 < P < 1
4 5 0 0.019 0.019 0 < P < 1
4 6 10 10.000 0.000 0 < P < 1
4 Cohort df= 2 1.975 P = 0.3724
5 1 8 8.311 0.012 0 < P < 1
5 2 3 2.125 0.360 0 < P < 1
5 3 0 0.438 0.438 0 < P < 1
5 4 0 0.110 0.110 0 < P < 1
5 5 0 0.016 0.016 0 < P < 1
5 6 54 54.000 0.000 0 < P < 1
5 Cohort df= 2 0.048 P = 0.9765
6 1 16 15.016 0.064 0 < P < 1
6 2 3 3.093 0.003 0 < P < 1
6 3 0 0.778 0.778 0 < P < 1
6 4 0 0.112 0.112 0 < P < 1
6 5 15 15.000 0.000 0 < P < 1
6 Cohort df= 2 0.307 P = 0.8575
7 1 25 25.290 0.003 0 < P < 1
7 2 3 5.210 0.937 0 < P < 1
7 3 3 1.310 2.179 0 < P < 1
7 4 1 0.189 3.468 0 < P < 1
7 5 198 198.000 0.000 0 < P < 1
7 Cohort df= 2 0.016 P = 0.9921
8 1 20 15.529 1.287 0 < P < 1
8 2 0 3.906 3.906 0 < P < 1
8 3 0 0.565 0.565 0 < P < 1
8 4 29 29.000 0.000 0 < P < 1
8 Cohort df= 2 5.758 P = 0.0562
9 1 38 40.376 0.140 0 < P < 1
9 2 13 10.155 0.797 0 < P < 1
9 3 1 1.468 0.149 0 < P < 1
9 4 203 203.000 0.000 0 < P < 1
9 Cohort df= 2 0.626 P = 0.7313
10 1 28 26.211 0.122 0 < P < 1
10 2 2 3.789 0.845 0 < P < 1
10 3 36 36.000 0.000 0 < P < 1
10 Cohort df= 2 0.967 P = 0.6166
11 1 39 40.189 0.035 0 < P < 1
11 2 7 5.811 0.243 0 < P < 1
11 3 210 210.000 0.000 0 < P < 1
11 Cohort df= 2 0.279 P = 0.8699
12 1 30 30.000 0.000 0 < P < 1
12 2 53 53.000 0.000 0 < P < 1
12 Cohort df= 1 0.000 P = 1.0000
13 1 46 46.000 0.000 0 < P < 1
13 2 194 194.000 0.000 0 < P < 1
13 Cohort df= 1 0.000 P = 1.0000
------------------------------------------------------------
@@ 4 0 0 30 41.8773 7 14.8486 -74.4986 186.997
G Total (Degrees of freedom = 30) 41.877
Pr(Larger Chi-square) = 0.0733
With pooling, Degrees of freedom = 7 Pearson Chi-square = 14.849
Pr(Larger Chi-square) = 0.0380
Log-likelihood = -74.498573 Akaike Information Criterion = 186.99715
CPU time in seconds for last procedure was 0.21
INPUT --- PROC TEST;
Submodel Name Log-likelihood NDF Akaike Inf. Criter. G-O-F
-------- ---------- -------------- --- ------------------- ------
1 SP -157.12123 47 318.24245 0.0000
2 SPG -105.87508 46 217.75016 0.0000
3 STPT -121.61377 36 269.22754 0.0000
4 STPTGT -74.498573 30 186.99715 0.0733
Likelihood Ratio Tests Between Models
General Reduced Degrees Pr(Larger
Submodel Submodel Chi-square Freedom Chi-square)
---------- ---------- ---------- ------- -----------
SPG SP 102.492 1 0.0000
STPT SP 71.015 11 0.0000
STPTGT SP 165.245 17 0.0000
STPT SPG 0.000 10 1.0000
STPTGT SPG 62.753 16 0.0000
STPTGT STPT 94.230 6 0.0000
* * WARNING * * Sequence of models reinitialized to zero.
CPU time in seconds for last procedure was 0.00
INPUT --- PROC STOP;
CPU time in minutes for this job was 0.01
E X E C U T I O N S U C C E S S F U L
Hardware Considerations
The program has also been compiled for DOS and Windows95.
Software Installation
To install TMSURVIV on a PC simply make a sub-directory to contain the programs and copy the files from the floppy disk. The disk contains the executable program file, so no compilation is necessary unless you wish to alter the dimensions. Here are the commands to install the TMSURVIV programs onto the hard disk of a PC:
c:> mkdir tmsrv
c:> cd tmsrv
c:> xcopy a:*.* /s
To install TMSURVIV on other computers, a FORTRAN compiler will be required. The files must first be transferred to disk on the computer, then compiled and linked into an executable program file. A "make" file is included which will create the executable program file from the source files if the target computer has the make utility (as most Unix systems do). If the target computer doesn't have a make utility, a "batch" file is included to compile all of the routines. Most likely, the make file or batch file will have to be edited to reflect the names of the compiler and linker on the target system.
Pradel, R., J.E. Hines, J-D. Lebreton, and J.D. Nichols. 1997. Capture-Recapture survival models taking account of transients. Biometrics 53, 60-72.
Pollock, K.H., J.D. Nichols, C. Brownie, and J.E. Hines. 1990. Statistical inference for capture-recapture experiments. Wildlife Monographs 107. 97pp.
White, G.C. 1983. Numerical estimation of survival rates from band-recovery and biotelemetry data. The Journal of Wildlife Management 47:716-728.
1. Formerly: National Biological Service