www.phidot.org

Hi all,

I have been running Burnham models in RMark, using the function approach described in paragraph C.13 to create and run a set of models in Mark and return the results in a marklist in Rmark.
Unfortunately, they tested the electrical installation at last weekend, resulting in a complete power shutdown.
At that time, Rmark/Mark had run about 70% of the models. Of these models I have the .inp, .res, .vcv and .out files, so I have the results of the models. But since the function in Rmark was not fully completed, the results are not returned as a marklist in Rmark.
I have tried to create a model list using collect.models, but this results in an error stating that no mark models can be found. I guess this means that although the models were successfully run in Mark, the results were not extracted from the output file and saved to a mark object in Rmark.

Is there a way to extract the results from the output files and to create a model list for the models I was able to run?
My best guess would be to use/adapt the internal function extract.mark.output?

Many thanks in advance!

This has arisen before and unless you use external=TRUE, I don't know of any way to do it easily. You would have to create each model with run =FALSE, set the matching Mark base filename and the extract the results. I think you woulf be better off re-running. One suggestion would be to run smaller sets and then merge the marklists. That would avoid losing too much work as long as you dod a save.image in between.Alternatively use external argument.

Thanks for your help.

Unfortunately I did not use external=TRUE, so I guess I will have to do it the hard way.
It took me about a month to run these models (they include individual covariates) so i would rather not run them again. For me it is probably quicker to just create all of the models with run=FALSE and then extract the results.

Next time I will definitely run them in batches and use the external argument. It would have saved me so much trouble. I guess you have to make mistakes before you can learn from them...