conda tidy verse installation issues

Question

[anthony.bertagnolli]

Greetings, i am trying to install the anvio-master branch. https://merenlab.org/2016/06/26/installation-v2/

I'm using a SLURM workload manager. Submission below:

#!/bin/bash
##first MaxBin run; MaxBin of random one fastq file
#SBATCH --job-name    condaTidyV        # job name
#SBATCH --output      condaTidyV-%j.out # standard output file (%j = jobid)
#SBATCH --error       condaTidyV-%j.err # standard error file
#SBATCH --partition   defq      # queue partition to run the job in
#SBATCH --nodes       1            # number of nodes to allocate
#SBATCH --ntasks-per-node 1        # number of descrete tasks - keep at one except for MPI 
#SBATCH --cpus-per-task=1          # number of CPU cores to allocate
#SBATCH --mem         20000         # 2000 MB of Memory allocated; set --mem with care
#SBATCH --time        24:00:00     # Maximum job run time
##SBATCH --mail-user   $email      # user to send emails to
##SBATCH --mail-type   ALL         # Email on: BEGIN, END, FAIL & REQUEUE
## Run 'man sbatch' for more information on the options above.

#module load python3

#conda config --add channels conda-forge

#conda config --set channel_priority strict

conda install -y -c bioconda r-tidyverse

The out file (condaTidyV-2175229.out) indicates it fails. I have tried several iterations of the installation.

Collecting package metadata (current_repodata.json): ...working... done
Solving environment: ...working... failed with initial frozen solve. Retrying with flexible solve.
Solving environment: ...working... failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): ...working... done
Solving environment: ...working... failed with initial frozen solve. Retrying with flexible solve.
Solving environment: ...working... 
Found conflicts! Looking for incompatible packages.
This can take several minutes.  Press CTRL-C to abort.
failed

Any help is welcome.

[ATpoint]

Please try with -c r rather than bioconda. See https://anaconda.org/r/r-tidyverse. And be sure to make a fresh and clean environment for the installation via conda create rather than pressing everything into the base environment, especially if it is such a big fish as the tidyverse. By the way, does your cluster policy really force you to submit things like simple compilations or conda installer jobs via SLURM rather than just run it on the head node? It is not a demanding jobs so you will not bring down the head with it.

[anthony.bertagnolli]

Thank you! Your suggestion worked. Much appreciated. I was using SLURM because my first attempts at installation took for ever (over an hour). I was afraid it might time out. I could increase the time allowed using a SLURM submission. Thanks again.

[ATpoint]

A good speed-up is also to use mamba rather than conda these days, see https://github.com/mamba-org/mamba

[ATpoint]

And for these things I always use screen on the headnode without the need to be logged in. On our HPC that is allowed (I guess...).