Entering edit mode
Hi,
I have just installed the new version or R (4.0.0) and Bioconductor 3.11, but I am getting the following warning every time I try to download a Bioconductor package.
> Bioconductor version 3.11 (BiocManager 1.30.10), R 4.0.0 (2020-04-24)
> Warning: unable to access index for repository
> https://bioconductor.org/packages/3.11/data/annotation/bin/macosx/contrib/4.0:
> Found continuation line starting ' script type="te ...' at begin
> of record. Warning: unable to access index for repository
> https://bioconductor.org/packages/3.11/data/experiment/bin/macosx/contrib/4.0:
> Found continuation line starting ' script type="te ...' at begin
> of record. Warning: unable to access index for repository
> https://bioconductor.org/packages/3.11/workflows/bin/macosx/contrib/4.0:
> Found continuation line starting ' script type="te ...' at begin
> of record.
sessionInfo()
R version 4.0.0 (2020-04-24)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Catalina 10.15.4
Matrix products: default
BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib
locale:
[1] en_GB.UTF-8/en_GB.UTF-8/en_GB.UTF-8/C/en_GB.UTF-8/en_GB.UTF-8
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] packrat_0.5.0 clustree_0.4.2 ggraph_2.0.2 ggplot2_3.3.0 Seurat_3.1.5 tibble_3.0.1 dplyr_0.8.5
loaded via a namespace (and not attached):
[1] viridis_0.5.1 httr_1.4.1 tidyr_1.0.3 tidygraph_1.1.2 jsonlite_1.6.1 viridisLite_0.3.0
[7] splines_4.0.0 lsei_1.2-0.1 leiden_0.3.3 assertthat_0.2.1 BiocManager_1.30.10 yaml_2.2.1
[13] ggrepel_0.8.2 globals_0.12.5 pillar_1.4.4 lattice_0.20-41 glue_1.4.0 reticulate_1.15
[19] digest_0.6.25 polyclip_1.10-0 RColorBrewer_1.1-2 colorspace_1.4-1 cowplot_1.0.0 htmltools_0.4.0
[25] Matrix_1.2-18 plyr_1.8.6 pkgconfig_2.0.3 tsne_0.1-3 listenv_0.8.0 purrr_0.3.4
[31] patchwork_1.0.0 scales_1.1.0 RANN_2.6.1 tweenr_1.0.1 Rtsne_0.15 ggforce_0.3.1
[37] farver_2.0.3 ellipsis_0.3.0 withr_2.2.0 ROCR_1.0-11 pbapply_1.4-2 lazyeval_0.2.2
[43] survival_3.1-12 magrittr_1.5 crayon_1.3.4 future_1.17.0 nlme_3.1-147 MASS_7.3-51.6
[49] ica_1.0-2 tools_4.0.0 fitdistrplus_1.0-14 data.table_1.12.8 lifecycle_0.2.0 stringr_1.4.0
[55] plotly_4.9.2.1 munsell_0.5.0 cluster_2.1.0 irlba_2.3.3 compiler_4.0.0 rsvd_1.0.3
[61] rlang_0.4.6 grid_4.0.0 ggridges_0.5.2 RcppAnnoy_0.0.16 htmlwidgets_1.5.1 igraph_1.2.5
[67] npsurv_0.4-0.1 gtable_0.3.0 codetools_0.2-16 graphlayouts_0.7.0 reshape2_1.4.4 R6_2.4.1
[73] gridExtra_2.3 zoo_1.8-8 knitr_1.28 uwot_0.1.8 future.apply_1.5.0 KernSmooth_2.23-17
[79] ape_5.3 stringi_1.4.6 parallel_4.0.0 Rcpp_1.0.4.6 vctrs_0.2.4 sctransform_0.2.1
[85] png_0.1-7 tidyselect_1.0.0 xfun_0.13 lmtest_0.9-37
How can I deal with this?
Many thanks, Lucy
Hi Lucy, are you using R Studio? There is a potential solution mentioned here:
For Bioconductor, to use 'http' rather than 'https' (NOT RECOMMENDED) via
options(BioC_mirror = "http://bioconductor.org")
Thank you both for the suggestions.
I have added
options(BioC_mirror = "http://bioconductor.org")
to my .Rprofile, however I am getting the same error just for http instead.I also tried unchecking the
Use secure download method for http
as suggested in the link provided by Kevin, but this didn't seem to solve the issue either.Can you try
and if that fails try visiting the value of
url
in a browser?I would guess that you are behind a 'proxy' and that you need to configure R to know about that. There are some instructions on
?download.file
@Martin: Lucy doesn't seem to have any problem with the PACKAGES index file of the software repo, only with the data annotation, data experiment, and workflows repos. These 3 repos are source only so these files are empty for the Mac binaries. In other words, with her setup
available.packages()
seems to have an issue with these empty PACKAGES files. No idea why.Ah... maybe it helps to see what
options("pkgType")
returns, perhaps updating tooptions(pkgType = "both")
. The output ofreadLines(url)
from the previous comment might also be useful, indicating what R thinks is at the other end of the url...After adding
options(pkgType = "both")
to my .Rprofile file, I still seem to be getting the warning messagesSince Hervé pointed out that the problem was not with the software repository, what about (taking the second repository...)
This gives:
How did you install R 4.0.0? The fact that you had
options("pkgType")
set tomac.binary
suggests that you didn't install the official CRAN binary. Note that compiling R from source or installing it via Homebrew or Fink is strongly discouraged and can lead to all kinds of problems. The authority on that matter, Simon Urbanek, needs to constantly remind people about this on the R-SIG-Mac mailing list. I don't know if that could be related to the warnings you're getting but I would start by making sure that you got that part correct.But your non-standard
options("pkgType")
setting could also mean that you have an.Rprofile
. Try to see if starting R with the--vanilla
option helps.Hi Herve,
I installed R by downloading the package from the CRAN website (https://cran.r-project.org). I definitely have an .Rprofile file that contains the following lines:
Could anything here be causing an issue? How would I open my project in RStudio with the equivalent of the
--vanilla
option?Many thanks,
Lucy
Just to note that I added the following two lines based on advice above:
That's the thing. If you had to add the
options(pkgType = "both")
line in order to get the standard type on Mac, that tells me that you probably didn't install the official R binary. With the standard R binary you get this set to"both"
out-of-the-box.Well we need to keep Rstudio out of the picture for now. Let's just focus on what happens with a bare R that you start at the command line by running
R --vanilla
in a terminal.Thank you - I just tried installing edgeR from Bioconductor in R (without the
--vanilla
option) and it worked fine, so it looks like it is something associated with my R project. Any ideas what this might be. My .Rprofile file is shown above.And in R on the command line,
options("pkgType")
is set to both.Can you get rid of (or comment out) the 1st line in your
.Rprofile
(options(packrat.dependency.discovery.disabled = TRUE)
) and try again? This time do NOT use the --vanilla option. If everything works fine (without the warning) then try again in RStudio.When I open R inside my project directory using
R --vanilla
, installing Bioconductor packages does not give an error.However, after removing the packrat line in my .Rprofile file, I still get the same error when trying to install packages inside RStudio.
Good. So everything works fine at the command line, with or w/o using
--vanilla
. Can you please try the following from within RStudio:and show us the output? Thanks