Creating input file for 3 occasions, 6 groups

[apryle]

I have hair samples of raccoons from 3 locations, male and female, which makes for 6 groups (male-loc1, fem-loc1, male-loc2, female-loc2, etc) in MARK. These data were collected during 3 sampling occasions. I am second-guessing my input file design and would appreciate confirmation as to whether or not the example below is correct. Specifically, I am wondering if an animal such as Rac1 below is seen as described,

ID occ sex location
Rac2 2 f 1
Rac3 1 f 2
Rac4 2 m 3
Rac1 1 m 1
Rac1 1 m 3
Rac1 2 m 2
Rac1 3 m 2

With the input file formatted such that:
Occ1Occ2Occ3 maleLoc1 femLoc1 maleLoc2 femaleLoc2 maleLoc3 femaleLoc3

My history should be (example1)...
010 0 1 0 0 0 0 /*Rac2*/
100 0 0 0 1 0 0 /*Rac3*/
010 0 0 0 0 1 0 /*Rac4*/
100 1 0 0 0 0 0 /*Rac1 was seen at loc1 on first occasion*/
100 0 0 0 0 1 0 /*Rac1 was seen at loc3 on first occasion*/
011 0 0 1 0 0 0 /*Rac1 was seen at loc2 on second and third occasion*/

You cannot have them belong to more than one group as shown in either of the cases below, correct (example2)?
010 0 1 0 0 0 0 /*Rac2*/
100 0 0 0 1 0 0 /*Rac3*/
010 0 0 0 0 1 0 /*Rac4*/
100 1 0 0 0 1 0 /*Rac1 seen at loc1 and loc3 on first occasion*/
010 0 0 1 0 0 0 /*Rac1 seen at loc2 on second occasion*/
001 0 0 1 0 0 0 /*Rac1 seen at loc2 on third occasion*/

And is incorrect because it states that Rac1 was seen at each sites at each of the occasions, correct (example3)?
010 0 1 0 0 0 0 /*Rac2*/
100 0 0 0 1 0 0 /*Rac3*/
010 0 0 0 0 1 0 /*Rac4*/
111 1 0 1 0 1 0 /*Rac1 seen at loc1, loc2, loc3, on all occasions*/

Thanks for clarification!

[simone77]

Hi,

The inp code for groups serves to make MARK knows that there is a certain number of groups in your data, like for instance if you have three groups:
1 0 0;
0 1 0;
0 0 1;

I am not sure about how MARK would interpret a code like this (but I don’t think it is a good idea):
1 0 0;
0 1 0;
0 0 1;
1 0 1;
That is what you are doing. If you thought that is biologically meaningful to split your data into a specific group for animals seen at site 1 and site 2 I would use four columns instead of three:
1 0 0 0;
0 1 0 0;
0 0 1 0;
0 0 0 1; # this being for animals seen at site 1 and 2.
However, without knowing your case study I would say that, a priori, it is an odd strategy to split your data base in such a way. No idea about your sample size except the number of occasions (3 is the very minimum one needs for running CR analyses), but I wonder if you might use a multi-site approach with 'A', 'B', and 'C' within the encounter history coding for each site. In that case you would have an extra parameter (besides survival) describing the probability one individual has of moving between one site and another.
Good luck!

[apryle]

Great, thank you for the confirmation. It sounds like the data in example1 is correct.