Quitting from lines ...

[nick-youngblut]

packrat::status() is quitting due to lines of code that I have in a .Rmd file. The error is:

> packrat::status()
Quitting from lines 80-92 (01_demultiplex.Rmd) 
Quitting from lines 80-92 (01_demultiplex.Rmd) 
Error in if (pkg$name == pkgName) return(pkg) : 
  argument is of length zero
In addition: There were 11 warnings (use warnings() to see them)

Lines 79-92 of 01_demultiplex.Rmd (one code chunk; R):

job = 'bash -c "source activate {conda_env}; {cmd}"'
cmd = "qiime tools import \
  --type EMPPairedEndSequences \
  --input-path {input_dir} \
  --output-path georg_seqs.qza"

cmd = sprintf(cmd, input_dir=georg_read_dir)

job = glue(job, conda_env=conda_env, cmd=cmd)

Sys.setenv(CMD=job)
print(job)

The multi-line string isn't the problem. If I comment-out these lines, they still cause the error. If I remove these lines, I get the following error:

> packrat::status()
running: bash  -c '# already ran
echo $CMD'
running: bash  -c 'ls -thlc *.qza'
Quitting from lines 90-91 (01_demultiplex.Rmd) 

The code chunk ({bash, cache=TRUE}) causing the problem is:

ls -thlc *.qza

Any ideas on what's going wrong? The code chunks don't seem that weird, so I would think that other users would have already run into similar problems.

sessionInfo()
R version 3.4.3 (2017-11-30)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 16.04.4 LTS

Matrix products: default
BLAS: /opt/microsoft/ropen/3.4.3/lib64/R/lib/libRblas.so
LAPACK: /opt/microsoft/ropen/3.4.3/lib64/R/lib/libRlapack.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C               LC_TIME=en_US.UTF-8       
 [4] LC_COLLATE=en_US.UTF-8     LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                  LC_ADDRESS=C              
[10] LC_TELEPHONE=C             LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] phyloseq_1.22.3      BiocInstaller_1.28.0 RevoUtilsMath_10.0.1

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.16        compiler_3.4.3      pillar_1.2.1        plyr_1.8.4         
 [5] XVector_0.18.0      iterators_1.0.9     tools_3.4.3         zlibbioc_1.24.0    
 [9] digest_0.6.15       packrat_0.4.9-1     evaluate_0.10.1     jsonlite_1.5       
[13] tibble_1.4.2        nlme_3.1-131        rhdf5_2.22.0        gtable_0.2.0       
[17] lattice_0.20-35     mgcv_1.8-22         pkgconfig_2.0.1     rlang_0.2.0        
[21] Matrix_1.2-12       foreach_1.4.4       igraph_1.2.1        yaml_2.1.18        
[25] parallel_3.4.3      knitr_1.20          stringr_1.3.0       cluster_2.0.6      
[29] Biostrings_2.46.0   RevoUtils_10.0.7    S4Vectors_0.16.0    IRanges_2.12.0     
[33] rprojroot_1.3-2     multtest_2.34.0     stats4_3.4.3        ade4_1.7-11        
[37] grid_3.4.3          Biobase_2.38.0      data.table_1.10.4-3 survival_2.41-3    
[41] rmarkdown_1.9       reshape2_1.4.3      ggplot2_2.2.1       magrittr_1.5       
[45] backports_1.1.2     htmltools_0.3.6     splines_3.4.3       scales_0.5.0       
[49] codetools_0.2-15    MASS_7.3-47         BiocGenerics_0.24.0 biomformat_1.6.0   
[53] permute_0.9-4       ape_5.1             colorspace_1.3-2    stringi_1.1.7      
[57] lazyeval_0.2.1      munsell_0.4.3       vegan_2.4-6 

[DaveParr]

Same issue. Also with sql chunks.

[joelnitta]

Same here:

Quitting from lines 149-156 (report.Rmd) 

The offending lines contain a ggplot2 call

ggplot(align_stats, aes(y=100*..count../sum(..count..), x=mean_coverage, color=mcl_settings)) +
  geom_density() +
  ylab("Rel. count (alignments)") +
  scale_y_continuous(labels = scales::percent) +
  xlab("mean non-missing data per alignment") +
  scale_x_continuous(labels = scales::percent)

And FWIW, I have another similar ggplot call soon after that doesn't cause any problems

ggplot(align_stats, aes(y=100*..count../sum(..count..), x=length, color=mcl_settings)) +
  geom_density() +
  ylab("Rel. count (alignments)") +
  scale_y_continuous(labels = scales::percent) +
  xlab("Alignment length (bp)")

[kevinushey]

A small, self-contained reproducible example would be helpful here. For example, a sample project (maybe hosted on GitHub) that, if I were to attempt to clone with Git and then call packrat::status() -- if I am able to reproduce the issue, then I can more easily investigate.

Attempting to call packrat::status() on a project containing a file hello.Rmd with the contents:

---
title: "Untitled"
output: html_document
---

```{r}
ggplot(align_stats, aes(y=100*..count../sum(..count..), x=mean_coverage, color=mcl_settings)) +
  geom_density() +
  ylab("Rel. count (alignments)") +
  scale_y_continuous(labels = scales::percent) +
  xlab("mean non-missing data per alignment") +
  scale_x_continuous(labels = scales::percent)
```

is not sufficient for me to reproduce the issue.

[joelnitta]

Sorry for the insufficient comment Kevin. I should have taken the time to document it properly, because now I can't get the error again either! I will post a proper reprex if it comes up again.

[mdpead]

I'm pretty sure this is a problem with the cache = TRUE option being set in the chunk. I had the same problem and removing this seems to have fixed it.

[dchiu911]

I am coming across this issue, and I think it has to do with packrat::status() not honouring the default knitr::opts_knit$set(eval.after = "fig.cap") when evaluating .Rmd files. I get two lines
Quitting from lines ... when setting fig.cap=cap with cap assigned in the chunk itself, but the warning disappears when I replace cap with the expression.

[dchiu911]

Hi @kevinushey

I've created a test project and ran a reprex at dchiu911/packrat-status#1. Could you try cloning to see if you can reproduce? Thanks!

[aronatkins]

This problem was caused by using rmarkdown::render to analyze R Markdown files for dependencies.

We believe that this is resolved with #647.