www.phidot.org

Hi all

I have a bug using export.MARK. The function does not succeed in creating the *.Rinp

For example : the multi-stata example in help give a .inp like that

Multistrata
1
4
0
3
A:Stratum1
B:Stratum2
C:Stratum3
1 1 1
1
group1
255
1
NA

This issues occurs under windows and linux.

Someone have an idea to fix this issue?

thanks in advance

Stephane

Huh, you are quite correct. I'll look into this and get back to you. --jeff

I take that back. I read your post quickly and got myself confused between the robust and mstrata examples. Actually all of them worked but you do have to manually remove the tmp files between examples or it will import the ones from say mallard and mstrata into robust,etc. I don't want to add the delete statements into the example because you have to use them for the import. Try running just the relevant parts of the example file after deleting all the .tmp files. Import and then delete the tmp files and then run the next part of they example.

They all imported fine for me. What version of RMark and MARK are you using? What error message are you getting?

--jeff

Hi Jeff

I was just trying the mstrata example, not all the examples.

My version of Mark and Rmark
R version 2.15.2 (2012-10-26)
This is RMark 2.1.4
I have mark for linux so I don't know the version (maybe old). But I have the same error on Windows (I have downloaded MARK for windwows today)

When I try to import the . Rinp in Mark (running under wine) I have the error : "Improper data type specified"

I also have this issue with my own multistrata models. The function return : NULL and just a .Rinp file
like that

MSLiveDead
1
31
0
2
b:Stratum1
G:Stratum2
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
7
fond_nlfond.age_lach0
fond_nlNe_libre.age_lach0
fond_nlfond.age_lach1
fond_nlfond.age_lach2
fond_nlfond.age_lach3
fond_nlfond.age_lach4
fond_nlfond.age_lach5
1061
1
NA

I will retry on windows and get back soon


PS : I want to export models into mark to do ANODEV analyses. So, I have a subsidiary question, I have not found a function in Rmark to do ANODEV. Is a such function implemented on RMark (with another name) ?

thanks for reply

stephane

There is no anodev function in RMark. I could probably write one if I had the time.

This error sounded familiar to me and see that there was a similar post on phidot in 2011 by bmiranda. Not sure what his situation was in regards to operating system. When you say it failed on Windows was this under Wine? Can you please try on a true Windows machine. It is working fine for me.

Unfortunately, this is not something I can help with because the error is occurring in the MARK interface. It is something that Gary would have to debug. Gary do you know what could be causing this error?

--jeff

It works perfectly on Windows native system. For the example and for my data.

I will retry later on linux (maybe it was just a problem of temporary file)

thanks for the help and sorry for the disturbance

stephane

scha wrote:It works perfectly on Windows native system. For the example and for my data.

I will retry later on linux (maybe it was just a problem of temporary file)

thanks for the help and sorry for the disturbance

stephane

Linux version is quite out of date at this point. Don't use it.

scha wrote:It works perfectly on Windows native system. For the example and for my data.

I will retry later on linux (maybe it was just a problem of temporary file)

thanks for the help and sorry for the disturbance

stephane

I suspect there is something buggy going on with the Linux MARK build as it is fairly old, I can not recreate your error on Windows on my end, but I did on one of the old Linux versions I had saved, so I think that is probably the issue.

bret

ok it should be the problem

I have got round the problem by doing an Anodev function in R

thank a lot for all your replies.

stephane

To scha ,

Would you mind sharing the R code for doing ANODEV tests in R? I am an R novice and it would be much appreciated!

Do these ANODEV tests include F-tests and p-values?