[R] Long jobs completing without output

Uwe Ligges ligges at statistik.tu-dortmund.de
Tue Dec 27 16:54:36 CET 2011



On 23.12.2011 14:54, Brendan Halpin wrote:
> I've been running a glmer logit on a very large data set (600k obs).
>
> Running on a 10% subset works correctly, but for the complete data set,
> R completes apparently without error, but does not display the results.
> Given these jobs take about 200 hours, it's very hard to make progress
> by trial and error.
>
> I append the code and the sample and complete output. As is apparent, I
> upgraded R during the complete run, but I recall testing on the
> subsample with the earlier version too. I am also assuming that
> upgrading R will not affect the running process -- is this true?

Err, this depends on the platform and the way you are using R.
If you change some parts of R while that is running, it may result into 
unexpected behaviour or a crash if R accesses the files after the 
upgrade, of course.

Best wishes,
Uwe Ligges



>
> I'd be grateful for any leads. In the meantime I'll be running with
> larger subsamples!
>
> Regards,
>
> Brendan Halpin
>
>
> - code ---------------------------------------------------------------
> library(arm)
> library(foreign)
> mlm<- read.dta("../workingdata.dta")
> attach(mlm)
>
> gender<- as.factor(stu_gend)
>
> yr<- year - 1998
> failure<- (lmer(fail ~
>                1 + cao + subj1 + subj2 + subj3 + gender + yr + ageentry + as.factor(yrs5)
>                  + modsize  + meancao + depfemr + (1|deptno) + (1|modinst)  + (1|ulid) ,
>                na.action = na.exclude, family = binomial (link="logit")))
>
> display(failure, digits=5, detail=TRUE)
> ----------------------------------------------------------------------
>
> - output with 10% sample data ----------------------------------------
> R version 2.14.0 (2011-10-31)
> Copyright (C) 2011 The R Foundation for Statistical Computing
> ISBN 3-900051-07-0
> Platform: i486-pc-linux-gnu (32-bit)
>
> R is free software and comes with ABSOLUTELY NO WARRANTY.
> You are welcome to redistribute it under certain conditions.
> Type 'license()' or 'licence()' for distribution details.
>
>    Natural language support but running in an English locale
>
> R is a collaborative project with many contributors.
> Type 'contributors()' for more information and
> 'citation()' on how to cite R or R packages in publications.
>
> Type 'demo()' for some demos, 'help()' for on-line help, or
> 'help.start()' for an HTML browser interface to help.
> Type 'q()' to quit R.
>
>> library(arm)
>
> arm (Version 1.4-13, built: 2011-6-19)
> Working directory is /home/brendan/work/mlmmarks/genderECSR
>> library(foreign)
>> mlm<- read.dta("../worksample-random1.dta")
>> attach(mlm)
>>
>> gender<- as.factor(stu_gend)
>>
>> yr<- year - 1998
>> failure<- (lmer(fail ~
> +               1 + cao + subj1 + subj2 + subj3 + gender + yr + ageentry + as.factor(yrs5)
> +                 + modsize  + meancao + depfemr + (1|deptno) + (1|modinst)  + (1|ulid) , na.action = na.exclude, family = binomial (link="logit")))
>>
>> display(failure, digits=5, detail=TRUE)
> glmer(formula = fail ~ 1 + cao + subj1 + subj2 + subj3 + gender +
>      yr + ageentry + as.factor(yrs5) + modsize + meancao + depfemr +
>      (1 | deptno) + (1 | modinst) + (1 | ulid), family = binomial(link = "logit"),
>      na.action = na.exclude)
>                   coef.est  coef.se   z value   Pr(>|z|)
> (Intercept)        2.63826   0.97870   2.69568   0.00702
> cao               -2.08963   0.11987 -17.43314   0.00000
> subj1              0.02608   0.23573   0.11064   0.91190
> subj2             -0.55668   0.32759  -1.69932   0.08926
> subj3             -1.57120   0.30664  -5.12400   0.00000
> genderM            0.36368   0.09188   3.95845   0.00008
> yr                 0.06067   0.01658   3.65996   0.00025
> ageentry          -0.00720   0.04338  -0.16598   0.86817
> as.factor(yrs5)1  -0.25181   0.05712  -4.40806   0.00001
> as.factor(yrs5)2  -0.54725   0.07601  -7.20005   0.00000
> as.factor(yrs5)3  -1.07483   0.08660 -12.41184   0.00000
> as.factor(yrs5)4  -1.22447   0.14373  -8.51932   0.00000
> as.factor(yrs5)5  -1.55032   0.31342  -4.94653   0.00000
> modsize            0.03387   0.02533   1.33733   0.18112
> meancao            1.08747   0.10748  10.11780   0.00000
> depfemr           -1.49097   0.49350  -3.02122   0.00252
>
> Error terms:
>   Groups   Name        Std.Dev.
>   modinst  (Intercept) 1.14308
>   ulid     (Intercept) 1.54030
>   deptno   (Intercept) 0.52497
>   Residual             1.00000
> ---
> number of obs: 63254, groups: modinst, 9076; ulid, 2275; deptno, 26
> AIC = 30275.2, DIC = 30237.2
> deviance = 30237.2
>>
> Loading required package: MASS
> Loading required package: Matrix
> Loading required package: lattice
>
> Attaching package: ‘Matrix’
>
> The following object(s) are masked from ‘package:base’:
>
>      det
>
> Loading required package: lme4
>
> Attaching package: ‘lme4’
>
> The following object(s) are masked from ‘package:stats’:
>
>      AIC, BIC
>
> Loading required package: R2WinBUGS
> Loading required package: coda
>
> Attaching package: ‘coda’
>
> The following object(s) are masked from ‘package:lme4’:
>
>      HPDinterval
>
> Loading required package: abind
> Loading required package: foreign
>
> Attaching package: ‘arm’
>
> The following object(s) are masked from ‘package:coda’:
>
>      traceplot
> ----------------------------------------------------------------------
>
> - output with complete data ------------------------------------------
> R version 2.13.1 (2011-07-08)
> Copyright (C) 2011 The R Foundation for Statistical Computing
> ISBN 3-900051-07-0
> Platform: i486-pc-linux-gnu (32-bit)
>
> R is free software and comes with ABSOLUTELY NO WARRANTY.
> You are welcome to redistribute it under certain conditions.
> Type 'license()' or 'licence()' for distribution details.
>
>    Natural language support but running in an English locale
>
> R is a collaborative project with many contributors.
> Type 'contributors()' for more information and
> 'citation()' on how to cite R or R packages in publications.
>
> Type 'demo()' for some demos, 'help()' for on-line help, or
> 'help.start()' for an HTML browser interface to help.
> Type 'q()' to quit R.
>
>> library(arm)
>
> arm (Version 1.4-13, built: 2011-6-19)
> Working directory is /home/brendan/work/mlmmarks/genderECSR
>> library(foreign)
>> mlm<- read.dta("../workingdata.dta")
>> attach(mlm)
>>
>> gender<- as.factor(stu_gend)
>>
>> yr<- year - 1998
>> failure<- (lmer(fail ~
> +               1 + cao + subj1 + subj2 + subj3 + gender + yr + ageentry + as.factor(yrs5)
> +                 + modsize  + meancao + depfemr + (1|deptno) + (1|modinst)  + (1|ulid) , na.action = na.exclude, family = binomial (link="logit")))
> Loading required package: MASS
> Loading required package: Matrix
> Loading required package: lattice
>
> Attaching package: ‘Matrix’
>
> The following object(s) are masked from ‘package:base’:
>
>      det
>
> Loading required package: lme4
>
> Attaching package: ‘lme4’
>
> The following object(s) are masked from ‘package:stats’:
>
>      AIC, BIC
>
> Loading required package: R2WinBUGS
> Loading required package: coda
>
> Attaching package: ‘coda’
>
> The following object(s) are masked from ‘package:lme4’:
>
>      HPDinterval
>
> Loading required package: abind
> Loading required package: foreign
>
> Attaching package: ‘arm’
>
> The following object(s) are masked from ‘package:coda’:
>
>      traceplot
> ----------------------------------------------------------------------
>



More information about the R-help mailing list