Error in eval(predvars, data, env) : object 'detect.obscured' not found

[klg-2016]

https://github.com/kaitlyngaynor/gorongosa-mesocarnivores/blob/ef9e7d850336f360b0bddf239328c83727af9791/scripts/02-multi-species-occupancy-analysis-GNP.Rmd#L188

Making this an issue to keep our conversation about it more in one place. I've now tried scaling the detect.obscured variable and it still throws this error.

[klg-2016]

Kaitlyn: I had that thought too, but it doesn't explain why it's working for me but not for you. I suppose you could try scaling it just to see—it won't affect your results, I think (AIC should stay the same) and maybe it's more appropriate anyway?

Truly don't know why. I have attached my "data" folder and you may try replacing yours with it (at least temporarily) to see if that fixes. That's one of the only differences between our setups since the code is identical. And presumably you are restarting R (which I am doing too) which means our packages and environment are the same.

The other difference is that I am running version 0.13-1 of unmarked. Could be the issue, but I don't want to update it in case it makes it break for me too! You might try to go back and reinstall an old version, I dunno.

[klg-2016]

I'll try with your data and a different version -- do you know how to reinstall an old version of the package?

[kaitlyngaynor]

Not off the top of my head but I've done it before, should be easily Googleable!

On Wed, Aug 19, 2020 at 11:53 AM klg-2016 notifications@github.com wrote:

I'll try with your data and a different version -- do you know how to reinstall an old version of the package?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/kaitlyngaynor/gorongosa-mesocarnivores/issues/73#issuecomment-676600477, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHA7WT3Q6DPIMQOWOLWYN7DSBQNTZANCNFSM4QFH5PCQ .

-- Kaitlyn Gaynor Postdoctoral researcher National Center for Ecological Analysis and Synthesis University of California - Santa Barbara

[klg-2016]

Okay, so I tried restarting R again, using your data folder, and installing the same version of unmarked and wasn't able to get it to run. Then I got code from a friend to use the workaround that Ken suggested in the google group, which now seems to be working! I just moved the detection covariates into obsCovs instead of siteCovs and now I can run predict().

[kaitlyngaynor]

Weird, but okay! Does it change the model summary or AIC or anything?

On Thu, Aug 20, 2020 at 8:24 AM klg-2016 notifications@github.com wrote:

Okay, so I tried restarting R again, using your data folder, and installing the same version of unmarked and wasn't able to get it to run. Then I got code from a friend to use the workaround that Ken suggested in the google group, which now seems to be working! I just moved the detection covariates into obsCovs instead of siteCovs and now I can run predict().

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kaitlyngaynor/gorongosa-mesocarnivores/issues/73#issuecomment-677732691, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHA7WT3BBQXFE4MHSSGSBLDSBU52NANCNFSM4QFH5PCQ .

-- Kaitlyn Gaynor Postdoctoral researcher National Center for Ecological Analysis and Synthesis University of California - Santa Barbara

[klg-2016]

I only checked the AIC, and it was off by 0.01. I didn't think to check the model summary. If you haven't pulled my changes and you have a min, would you mind sending me a screenshot of the model summary for one of the models and I'll check it against what I run now?

[klg-2016]

(if not, no problem and I can just undo it and save the summary, and then redo)

[kaitlyngaynor]

sure, which model?

On Thu, Aug 20, 2020 at 9:11 AM klg-2016 notifications@github.com wrote:

(if not, no problem and I can just undo it and save the summary, and then redo)

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kaitlyngaynor/gorongosa-mesocarnivores/issues/73#issuecomment-677759012, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHA7WTY5SP5PXQDGNNBO3VDSBVDJZANCNFSM4QFH5PCQ .

-- Kaitlyn Gaynor Postdoctoral researcher National Center for Ecological Analysis and Synthesis University of California - Santa Barbara

[klg-2016]

thank you! could you run fit3?

[kaitlyngaynor]

here is fit 3

[image: Screen Shot 2020-08-20 at 9.15.41 AM.png]

On Thu, Aug 20, 2020 at 9:13 AM Kaitlyn Gaynor gaynor@nceas.ucsb.edu wrote:

sure, which model?

On Thu, Aug 20, 2020 at 9:11 AM klg-2016 notifications@github.com wrote:

(if not, no problem and I can just undo it and save the summary, and then redo)

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kaitlyngaynor/gorongosa-mesocarnivores/issues/73#issuecomment-677759012, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHA7WTY5SP5PXQDGNNBO3VDSBVDJZANCNFSM4QFH5PCQ .

-- Kaitlyn Gaynor Postdoctoral researcher National Center for Ecological Analysis and Synthesis University of California - Santa Barbara

-- Kaitlyn Gaynor Postdoctoral researcher National Center for Ecological Analysis and Synthesis University of California - Santa Barbara

[kaitlyngaynor]

Oops maybe that didn't work, I sent by e-mail. Trying again here.

Also, did you already try scaling it? That didn't work either? I'm just puzzled!

[klg-2016]

That second try worked, thank you!

Try scaling it, you mean scaling the covariates?

[kaitlyngaynor]

Yeah, scaling the detection covariate (we had discussed that as a potential issue, since it was the only variable that hadn't been scaled)

On Thu, Aug 20, 2020 at 9:18 AM klg-2016 notifications@github.com wrote:

That second try worked, thank you!

Try scaling it, you mean scaling the covariates?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kaitlyngaynor/gorongosa-mesocarnivores/issues/73#issuecomment-677763146, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHA7WT3NYOQ6RGIOL62RNQDSBVEGRANCNFSM4QFH5PCQ .

-- Kaitlyn Gaynor Postdoctoral researcher National Center for Ecological Analysis and Synthesis University of California - Santa Barbara

[klg-2016]

They're very close but not quite identical:

Yep, I did try scaling detect.obscured and it didn't address the error

[kaitlyngaynor]

Ok, I think close enough. If the person who wrote the function suggested this as a workaround, should be good enough for us! I just don't like that it works on my computer and not on yours and that we don't know why, haha

On Thu, Aug 20, 2020 at 9:21 AM klg-2016 notifications@github.com wrote:

They're very close but not quite identical: [image: image] https://user-images.githubusercontent.com/66888714/90798138-8bd81500-e2df-11ea-88d2-da7c7fafd8d5.png

Yep, I did try scaling detect.obscured and it didn't address the error

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kaitlyngaynor/gorongosa-mesocarnivores/issues/73#issuecomment-677764564, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHA7WT37TZC3R2A5SAM7QVDSBVEP5ANCNFSM4QFH5PCQ .

-- Kaitlyn Gaynor Postdoctoral researcher National Center for Ecological Analysis and Synthesis University of California - Santa Barbara

[klg-2016]

No, absolutely me either. I explained all that to my brother last night, and he was going to see if he could get it to run on a Mac he has with the same version you're using, and maybe get somewhere with understanding why there's a difference (which maybe is just the PC vs Mac, but even that shouldn't be happening). I'll let you know if we get anywhere with it.

Also, unrelated to this issue: can I share the code you wrote to create these graphs (the one in the screenshot) with my friend who's also using this model?

[kaitlyngaynor]

Whoooo knows. Another thing is that I have been way too scared to upgrade to R 4.0 because I have heard it has messed up some packages, so I'm still running R version 3.6.3 (2020-02-29), the last version before 4.0 came out in June. (you can look at R.Version() to see the difference). So maybe that could be it too? Again, I have no idea, but it seems like it doesn't matter terribly much.

And yes, please share code!!!

On Thu, Aug 20, 2020 at 9:28 AM klg-2016 notifications@github.com wrote:

No, absolutely me either. I explained all that to my brother last night, and he was going to see if he could get it to run on a Mac he has with the same version you're using, and maybe get somewhere with understanding why there's a difference (which maybe is just the PC vs Mac, but even that shouldn't be happening). I'll let you know if we get anywhere with it.

Also, unrelated to this issue: can I share the code you wrote to create these graphs (the one in the screenshot) with my friend who's also using this model?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/kaitlyngaynor/gorongosa-mesocarnivores/issues/73#issuecomment-677768288, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHA7WTZIPIRVMN6LBCELQEDSBVFKLANCNFSM4QFH5PCQ .

-- Kaitlyn Gaynor Postdoctoral researcher National Center for Ecological Analysis and Synthesis University of California - Santa Barbara

[klg-2016]

Hahah I'm running an even older version than you apparently: R version 3.6.1 (2019-07-05). I could say that was for the same reason you just said, but I just didn't realize I should upgrade (even though we had a very brief R session this spring, and I'm surprised that didn't come up as a recommendation).

Thank you thank you!