R script stops running when I am not actively using the computer

Question

I'm running a script that performs automated backwards selection on glmer models. The script runs sequential drop1() commands, automatically removing the predictor with the highest p-value after each step. It's very time-consuming so I want to leave it to run overnight, but R stops running when I leave the computer.

I'm running R 3.1.1 in Snow Leopard. I have already set the System Preferences to prevent the computer from sleeping, and I am also preventing sleep by running "pmset noidle" in Terminal. Despite all this, when I leave my script to run, it silently pauses about 45 minutes after my last activity (in any application), and silently resumes when I start using the computer again in the morning. I can tell approximately when the script is active because it periodically prints timestamps. I've pasted an example fragment of the output at the bottom of this post (mostly so folks can see the timestamps). I recognize that I'm getting some warnings related to the model fitting, but I'm pretty sure these are not causing the 7.5-hour delay!

Has this happened to anyone else? Any suggestions for how to keep my script running while I'm away from the computer? I'm new to the site, so please let me know what additional info I can provide.

[1] "Wed Feb 25 21:31:59 2015"
Single term deletions

Model:
Pref ~ Nut * Subfamily + Nut * Sex + Subfamily * Sex + DrinkDur + 
    TrialTime + holdDur + FeedOrder + (1 | SpCode/Indiv)
              Df    AIC    LRT   Pr(Chi)    
<none>           809.07                     
DrinkDur       1 810.77  3.703   0.05433 .  
TrialTime      1 807.06 -0.011   1.00000    
holdDur        1 807.37  0.306   0.58036    
FeedOrder      1 810.24  3.172   0.07490 .  
Nut:Subfamily 32 820.47 75.404 2.315e-05 ***
Nut:Sex        4 843.44 42.375 1.395e-08 ***
Subfamily:Sex  8 809.06 15.995   0.04245 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

[1] "Now dropping TrialTime"
[1] "Wed Feb 25 22:39:51 2015"
Single term deletions

Model:
Pref ~ Nut + Subfamily + Sex + DrinkDur + holdDur + FeedOrder + 
    (1 | SpCode/Indiv) + Nut:Subfamily + Nut:Sex + Subfamily:Sex
              Df    AIC    LRT   Pr(Chi)    
<none>           807.06                     
DrinkDur       1 808.72  3.665   0.05556 .  
holdDur        1 805.33  0.276   0.59950    
FeedOrder      1 808.19  3.134   0.07667 .  
Nut:Subfamily 32 818.18 75.125 2.523e-05 ***
Nut:Sex        4 841.42 42.360 1.405e-08 ***
Subfamily:Sex  8 807.18 16.125   0.04062 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
[1] "Now dropping holdDur"
[1] "Thu Feb 26 00:39:02 2015"
Single term deletions

Model:
Pref ~ Nut + Subfamily + Sex + DrinkDur + FeedOrder + (1 | SpCode/Indiv) + 
    Nut:Subfamily + Nut:Sex + Subfamily:Sex
              Df    AIC    LRT   Pr(Chi)    
<none>           805.33                     
DrinkDur       1 806.85  3.519   0.06066 .  
FeedOrder      1 806.47  3.134   0.07667 .  
Nut:Subfamily 32 816.48 75.144 2.508e-05 ***
Nut:Sex        4 839.69 42.357 1.407e-08 ***
Subfamily:Sex  8 805.55 16.220   0.03934 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
[1] "Now dropping FeedOrder"
[1] "Thu Feb 26 08:04:21 2015"
There were 50 or more warnings (use warnings() to see the first 50)
Single term deletions

Model:
Pref ~ Nut + Subfamily + Sex + DrinkDur + (1 | SpCode/Indiv) + 
    Nut:Subfamily + Nut:Sex + Subfamily:Sex
              Df    AIC    LRT   Pr(Chi)    
<none>           806.47                     
DrinkDur       1 808.10  3.636   0.05655 .  
Nut:Subfamily 32 819.11 76.647 1.574e-05 ***
Nut:Sex        4 840.14 41.678 1.945e-08 ***
Subfamily:Sex  8 806.69 16.222   0.03931 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
[1] "Now dropping DrinkDur"
[1] "Thu Feb 26 09:48:59 2015"
Single term deletions

Model:
Pref ~ Nut + Subfamily + Sex + (1 | SpCode/Indiv) + Nut:Subfamily + 
    Nut:Sex + Subfamily:Sex
              Df    AIC    LRT   Pr(Chi)    
<none>           808.10                     
Nut:Subfamily 32 821.02 76.918 1.446e-05 ***
Nut:Sex        4 842.05 41.944 1.713e-08 ***
Subfamily:Sex  8 807.25 15.147   0.05635 .  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
[1] "Now dropping Subfamily:Sex"
[1] "Thu Feb 26 10:02:58 2015"
Single term deletions

Model:
Pref ~ Nut + Subfamily + Sex + (1 | SpCode/Indiv) + Nut:Subfamily + 
    Nut:Sex
              Df    AIC    LRT   Pr(Chi)    
<none>           807.25                     
Nut:Subfamily 32 829.94 86.695 6.160e-07 ***
Nut:Sex        4 849.83 50.576 2.737e-10 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Answer 1

I'm not sure why you're having this problem, given the options that you've used to prevent the computer from going to sleep (did you also disable the screensaver?).

Let's just work around the problem instead of trying to figure out why your computer is entering a sleep or idle state despite your attempts to prevent this with settings.

Here's what you should do:

1. Open a terminal window (you can find it in Spotlight)
2. type screen to enable the multiplexing screen program, which is useful for persistence (used mainly when connecting remotely to Linux servers, but also useful here)
3. Run your script in the terminal window, editing the following command for your specific R installation and script location: /usr/bin/Rscript "/Users/usr/Documents/longscript.R"

That should do the trick. If you want to be extra cautious or if you have any trouble, try running Caffeinate first.