Unusual encounter history problem

[cooch]

My point is it appears that when there are more encounters than individuals it is stopping me fit models. When I try to fit the model is when I get the

"Error - encounter history was too short for the number of occasions specified."

And MARK crashes.

Then that points to a problem either with your input file, how you're specifying the problem when setting up the data type, or both.

[HenSH]

That was my original issue. I don't believe it is an issue with the input file. If I have 62 individuals and 64 encounters it does not work. If I delete the first 3 encounters it works perfectly. It have tried this with numerous datasets to test this and it is always the same.

Thanks so much for trying to help me with this issue.

[cooch]

That was my original issue. I don't believe it is an issue with the input file. If I have 62 individuals and 64 encounters it does not work. If I delete the first 3 encounters it works perfectly. It have tried this with numerous datasets to test this and it is always the same.

Thanks so much for trying to help me with this issue.

Nothing more I can do. 99% of the time, such problems start with something at the 'user' end -- either the .inp file, or the data type specification.

[mpkeating]

Hi all, first time poster long time lurker! Curious to know if this issue was ever resolved or if anyone has had better luck with RMark? I am also running into this issue where I can’t fit models when my encounter occasions > entries (ie, semicolons). I am currently experiencing internet outages in the western Carolinas so if this was resolved but I haven’t found it, I apologize in advance and would be grateful for a link to the thread!

To be clear, I’ve used MARK many times and am certain there is no issue with my .inp file. In fact, when I enter the same file with number of occasions </= entries I have no issues - but of course can’t fit models appropriately. As some added info, I’m looking at 99 individual entries and 365 encounter occasions. Thanks!

[jhines]

I tested Mark with simulated data with 100 individuals and 365 occasions and the CJS model. I got the warning about sparse data, but was able to run models without error. I did get Mark to crash by entering the wrong number of occasions, so that might be something to check. If you're still having trouble, perhaps you could post the first few lines of input or send me (jhines@usgs.gov) the entire input file off-list.

Jim

[jlaake]

Have seen this posted before but never saw anyone explain where the errors are occurring. Nor did anyone send or offer to send a file that demonstrated the problem. Without that any solution is just a guess. MARK .inp files need to be converted to the RMark format to be processed. When you run a model with the mark function, it creates a .inp file in the format for MARK to read. If there is a problem it would be at that point so it is that .inp file that would have to be examined and the accompanying .out file for any error messages. With a large number of occasions, the capture histories may be getting wrapped across multiple lines and maybe that could be the problem if that format doesn't agree with MARK. Just a guess. If Jim can't figure it out he can send me the file and I'll look at it. Thanks Jim.

--jeff