kundajelab / atac_dnase_pipelines

ATAC-seq and DNase-seq processing pipeline
BSD 3-Clause "New" or "Revised" License
161 stars 81 forks source link

Install error during IDR replacement section of install_dependencies.sh: unable to execute 'gcc' #90

Closed AmandaBirmingham closed 6 years ago

AmandaBirmingham commented 6 years ago

When attempting to install the pipeline on a new, fresh linux instance from Amazon Web Services (I have tried on both Amazon Linux 2017.09.1 and Ubuntu 16.04 instances), I encounter a fatal error while running install_dependencies.sh. The error occurs during the section of the install script that uninstalls IDR 2.0.3 and installs the latest one; I think it occurs during the 'python3 setup.py install' call. Here are the messages I receive:

gcc -pthread -Wsign-compare -Wunreachable-code -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I/home/ubuntu/miniconda3/envs/bds_atac_py3/lib/python3.5/site-packages/numpy/core/include -I/home/ubuntu/miniconda3/envs/bds_atac_py3/include/python3.5m -c idr/inv_cdf.c -o build/temp.linux-x86_64-3.5/idr/inv_cdf.o unable to execute 'gcc': No such file or directory error: command 'gcc' failed with exit status 1

As you can see, the conda environment appears to be failing to find a gcc compiler. Perhaps this is related to a reported conda issue (https://github.com/conda/conda/issues/6030 ) but this issue appears to have been fixed in the conda version that the install is using (running conda -V gives me 'conda 4.4.7').

Below are the commands that I ran on a fresh Ubuntu 16.04 instance from AWS that reproduce this error: sudo add-apt-repository ppa:webupd8team/java -y sudo apt-get update sudo apt-get install oracle-java8-installer wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh bash Miniconda3-latest-Linux-x86_64.sh [log out, log back in to re-source] wget https://github.com/leepc12/BigDataScript/blob/master/distro/bds_Linux.tgz?raw=true -O bds_Linux.tgz mv bds_Linux.tgz $HOME && cd $HOME && tar zxvf bds_Linux.tgz [get msg: mv: 'bds_Linux.tgz' and '/home/ubuntu/bds_Linux.tgz' are the same file] nano $HOME/.bashrc [add export PATH=$PATH:$HOME/.bds to end of bashrc, then write out and exit nano] [log out, log back in to re-source] git clone https://github.com/kundajelab/atac_dnase_pipelines --recursive cd atac_dnase_pipelines [Note that this command is NOT in install instructions, but is necessary to get to the directory where the next command can be recognized] bash install_dependencies.sh

Any help resolving this issue would be greatly appreciated!

leepc12 commented 6 years ago

I am sorry for late response.

Can you try with Conda 4.4.3 too? You can find it on https://repo.continuum.io/miniconda/ https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh. That's the version I am using.

Also please make sure that you have gcc installed on your system. It's weird why the very basic compiler like gcc doesn't exist on your system. I also use Ubuntu 16.04 but gcc is pre-installed when OS is installed.

sudo apt-get install gcc gfortran

Thanks,

Jin

On Thu, Jan 11, 2018 at 10:25 AM, Amanda Birmingham < notifications@github.com> wrote:

When attempting to install the pipeline on a new, fresh linux instance from Amazon Web Services (I have tried on both Amazon Linux 2017.09.1 and Ubuntu 16.04 instances), I encounter a fatal error while running install_dependencies.sh. The error occurs during the section of the install script that uninstalls IDR 2.0.3 and installs the latest one; I think it occurs during the 'python3 setup.py install' call. Here are the messages I receive:

gcc -pthread -Wsign-compare -Wunreachable-code -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I/home/ubuntu/miniconda3/ envs/bds_atac_py3/lib/python3.5/site-packages/numpy/core/include -I/home/ubuntu/miniconda3/envs/bds_atac_py3/include/python3.5m -c idr/inv_cdf.c -o build/temp.linux-x86_64-3.5/idr/inv_cdf.o unable to execute 'gcc': No such file or directory error: command 'gcc' failed with exit status 1

As you can see, the conda environment appears to be failing to find a gcc compiler. Perhaps this is related to a reported conda issue ( conda/conda#6030 https://github.com/conda/conda/issues/6030 ) but this issue appears to have been fixed in the conda version that the install is using (running conda -V gives me 'conda 4.4.7').

Below are the commands that I ran on a fresh Ubuntu 16.04 instance from AWS that reproduce this error: sudo add-apt-repository ppa:webupd8team/java -y sudo apt-get update sudo apt-get install oracle-java8-installer wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh bash Miniconda3-latest-Linux-x86_64.sh [log out, log back in to re-source] wget https://github.com/leepc12/BigDataScript/blob/master/ distro/bds_Linux.tgz?raw=true -O bds_Linux.tgz mv bds_Linux.tgz $HOME && cd $HOME && tar zxvf bds_Linux.tgz [get msg: mv: 'bds_Linux.tgz' and '/home/ubuntu/bds_Linux.tgz' are the same file] nano $HOME/.bashrc [add export PATH=$PATH:$HOME/.bds to end of bashrc, then write out and exit nano] [log out, log back in to re-source] git clone https://github.com/kundajelab/atac_dnase_pipelines --recursive cd atac_dnase_pipelines [Note that this command is NOT in install instructions, but is necessary to get to the directory where the next command can be recognized] bash install_dependencies.sh

Any help resolving this issue would be greatly appreciated!

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/kundajelab/atac_dnase_pipelines/issues/90, or mute the thread https://github.com/notifications/unsubscribe-auth/AIOd_Abg5uyNMIU8gaNvel-APIUW7iaiks5tJlI0gaJpZM4RbQI3 .

leepc12 commented 6 years ago

closing this due to long inactivity