Installation notes

This section lists software which serve as backbone for a variety of projects including those in genetics. Illustration is given for some under Ubutun except R-devel which is with Fedora whose C/C++ version is higher.

Environment modules

Web: https://modules.readthedocs.io/en/latest/

It is preferable to allow for installation of multiple applications. The following scripts show how this is done under Ubunto.

wget http://archive.ubuntu.com/ubuntu/pool/universe/m/modules/modules_5.2.0.orig.tar.xz
tar xf modules_5.2.0.orig.tar.xz
cd modules-5.2.0
sudo apt-get install tcl-dev tk-dev
./configure
make
sudo make install
ls /usr/local/Modules
source /usr/local/Modules/init/bash
sudo ln -s /usr/local/Modules/init/profile.sh /etc/profile.d/modules.sh
sudo ln -s /usr/local/Modules/init/profile.csh /etc/profile.d/modules.csh
echo -e "\n# For initiating Modules" | sudo tee -a /etc/bash.bashrc > /dev/null  # Append a line to the end of this file with no return message.
echo ". /etc/profile.d/modules.sh" | sudo tee -a /etc/bash.bashrc > /dev/null
less /usr/local/Modules/init/profile.sh
module avail
module list

According to https://www.microbialsystems.cn/en/post/xubuntu_env_modules/. Instances at work is shown here, https://cambridge-ceu.github.io/csd3/systems/ceuadmin.html.

Armadillo

It is available with

sudo apt install libarmadillo-dev

boost

It is installed with

sudo apt install libboost-all-dev

To install it manually from source, as for a particular version, https://stackoverflow.com/questions/12578499/how-to-install-boost-on-ubuntu

wget https://sourceforge.net/projects/boost/files/boost/1.58.0/boost_1_58_0.tar.gz
tar xvfz boost_1_58_0.tar.gz
cd boost_1_58_0
# ./b2 -h gives more options
./bootstrap.sh --prefix=/scratch/jhz22
./b2

With a successful built, the following directory is suggested to be added to compiler include paths:

boost_1_58_0

The following directory should be added to linker library paths:

boost_1_58_0/stage/lib

and we can test with example

#include <iostream>
#include <boost/array.hpp>

using namespace std;
int main(){
  boost::array<int, 4> arr = {{1,2,3,4}};
  cout << "hi" << arr[0];
  return 0;
}

eigen

It is installed with

sudo apt install libeigen3-dev

GMP/MPFR

One can start usual from https://gmplib.org/ and https://www.mpfr.org/.

sudo apt install libgmp-dev
sudo apt install libmpfr-dev

then one can install Rmpfr. When installing as non-Admin, make sure issuing 'make check' for both libraries. As MPFR is dependent on GMP, it is necessary to use

cd /home/jhz22/Downloads/mpfr-4.0.1
./configure --prefix=/scratch/jhz22 --with-gmp-build=/home/jhz22/Downloads/gmp-6.1.2
make check

for instance.

GSL

sudo apt install libgsl-dev

JAGS-4.3.0

These are required at least under Federa 28,

sudo dnf install automake
sudo dnf install lapack-devel
sudo dnf install mercurial

It is actually available from Ubuntu archive, i.e.,

sudo apt install jags
sudo apt-get install r-cran-rjags

We can also work with sourceforge,

wget https://sourceforge.net/projects/mcmc-jags/files/JAGS/4.x/Source/JAGS-4.3.0.tar.gz
tar xvfz JAGS-4.3.0.tar.gz
cd JAGS-4.3.0
LDFLAGS="-L/scratch/jhz22/lib64" ./configure --prefix=/scratch/jhz22 --with-blas=-lblas --with-lapack=-llapack
make
make install

Under MKL, we have

#22-7-2014 MRC-Epid JHZ

export MKL_NUM_THREAD=15
export MKL=/home/jhz22/intel/composer_xe_2013.4.183/mkl
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$MKLROOT/lib/intel64
/genetics/data/software/intel/composer_xe_2013.4.183/mkl/bin/mklvars.sh intel64
./configure --prefix=/home/jhz22 --disable-shared --with-lapack \
--with-blas="-fopenmp -m64 -I$MKL/include -L$MKL/lib/intel64 -lmkl_gf_lp64 -lmkl_gnu_thread -lmkl_core -lpthread -lm"
make

It turns out the easiest to install rjags package is to download it and work manually, e.g.,

R --no-save <<END
  download.packages("rjags",".",repos="https://cran.r-project.org")
END
tar xvfz rjags_4-8.tar.gz
cd rjags
rm configure
cd src
(
  echo CPP_FLAGS=-I/usr/local/include/JAGS
  echo LDFLAGS=-L/usr/local/lib
) > Makevars
cd ../..
R CMD INSTALL rjags

The rjags package can also be installed as follows,

export PKG_CONFIG_PATH=/scratch/jhz22/lib/pkgconfig

R CMD INSTALL rjags_4-6.tar.gz --configure-args='CPPFLAGS="-fPIC" LDFLAGS="-L/scratch/jhz22/lib -ljags"
--with-jags-prefix=/scratch/jhz22
--with-jags-libdir=/scratch/jhz22/lib
--with-jags-includedir=/scratch/jhz22/include'

It may still be difficult to install, and we can try manually,

tar xfz rjags_4-6.tar.gz
cd rjags
mv configure configure.bak
echo PKG_CPPFLAGS=-fPIC -I/scratch/$USER/include/JAGS > src/Makevars
echo PKG_LIBS=-L/scratch/$USER/lib -ljags >> src/Makevars
cd -
R CMD INSTALL rjags

After this, rjags should install as with R2jags.

We can also install JAGS-related packages by establishing an Makevars in the src directory, e.g.,

R --no-save <<END
download.packages("runjags",".")
END
tar xvfz runjags_2.0.4-2.tar.gz
cd runjags
mv configure configure.bak
# modify jagsversions.h
# cp Makevars.runjags src/Makevars
cd -
R CMD INSTALL runjags

We need to add modify runjags/src/jagsversions.h to proceed,

#ifndef JAGS_VERSIONS_H_
#define JAGS_VERSIONS_H_

#include <version.h>
#ifndef JAGS_MAJOR
#define JAGS_MAJOR 4
#endif

#define JAGS_MAJOR_FORCED 0

where the Makevars.runjags has the following lines

PKG_CPPFLAGS=-I/scratch/jhz22/include
PKG_LIBS=-L/scratch/jhz22/lib -ljags 

OBJECTS= distributions/jags/DFunction.o distributions/jags/DPQFunction.o distributions/jags/PFunction.o distributions/jags/QFunction.o distributions/jags/RScalarDist.o distributions/DPar1.o distributions/DPar2.o distributions/DPar3.o distributions/DPar4.o distributions/DLomax.o distributions/DMouchel.o distributions/DGenPar.o distributions/DHalfCauchy.o runjags.o testrunjags.o

To get around these, one can mirror installation of rjags using the fact that runjags simply calls libjags.so though the source seemed for JAGS 3.x.x.,

export PKG_CONFIG_PATH=/rds-d4/user/jhz22/hpc-work/lib/pkgconfig
export LDFLAGS="-L/rds-d4/user/jhz22/hpc-work/lib -ljags -lblas -llapack"
R CMD INSTALL runjags_2.0.4-2.tar.gz --configure-args='
--with-jags-prefix=/rds-d4/user/jhz22/hpc-work
--with-jags-libdir=/rds-d4/user/jhz22/hpc-work/lib
--with-jags-includedir=/rds-d4/user/jhz22/hpc-work/include'

but somehow runjags is always points to lib64 for libjags.so, so when libjags.so is in lib instead it is necessary to create symbolic links from lib64.

BLAS and LAPACK

The pre-built version is straightforward for Fedora with

sudo dnf install blas-devel
sudo dnf install lapack-devel

and the counterpart for Ubuntu is

sudo apt install libblas-dev
sudo apt install liblapack-dev

To install from http://www.netlib.org/lapack/, we proceed as follows,

wget http://www.netlib.org/lapack/lapack-3.8.0.tar.gz
tar xvfz lapack-3.8.0.tar.gz
cd lapack-3.8.0
mkdir build
cd build
## ccmake .
cmake ..
make
make install

It is necessary to invoke ccmake .. to change the default static to dyanmic library as well as target directory. However, in case this is working, one can proceed as follows,

cmake -DCMAKE_INSTALL_PREFIX=/rds-d4/user/jhz22/hpc-work -DCMAKE_BUILD_TYPE=RELEASE -DBUILD_SHARED_LIBS=ON -DCBLAS=ON -DLAPACKE=ON ..
make
make install

MKL

One can consult Intel® Math Kernel Library Link Line Advisor and Free access to Intel® Compilers, Performance libraries, Analysis tools and more....

For instance, it is conviently available from Anaconda,

conda install -c intel mkl

Example use with R under RHEL,

# export OMP_NUM_THREADS=6
export MKL_NUM_THREADS=15
export MKLROOT=/genetics/data/software/intel/composer_xe_2013.4.183/mkl
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$MKLROOT/lib/intel64
/genetics/data/software/intel/composer_xe_2013.4.183/mkl/bin/mklvars.sh intel64
./configure --prefix=/genetics/data/software --enable-R-shlib --enable-threads=posix --with-lapack \
 --with-blas="-fopenmp -m64 -I$MKLROOT/include -L$MKLROOT/lib/intel64 -lmkl_gf_lp64 -lmkl_gnu_thread -lmkl_core -lpthread -lm"
make
make install

and

# https://software.intel.com/en-us/articles/build-r-301-with-intel-c-compiler-and-intel-mkl-on-linux#
export ICC_OPT="-mkl -xHOST -fp-model strict"
export CC="icc $ICC_OPT"
export CXX="icpc $ICC_OPT"
export FC="ifort -mkl -xHOST"
export F77="ifort -mkl -xHOST"
export FPICFLAGS=" -fPIC"
export AR=xiar
export LD=xild
export MKL="-lmkl_gf_lp64 -lmkl_intel_thread  -lmkl_core -liomp5 -lpthread"
./configure --prefix=/home/jhz22/R-devel --enable-R-shlib --with-x=no --with-blas=-lmkl LDFLAGS=-L/home/jhz22/lib CPPFLAGS=-I/home/jhz22/include

For Windows, see https://software.intel.com/content/www/us/en/develop/documentation/get-started-with-mkl-for-windows/top.html. The benchmark is available from here, https://github.com/pachamaltese/r-with-intel-mkl/blob/master/00-benchmark-scripts/1-r-benchmark-25.R.

cd "C:\Program Files\R\R-4.0.0\bin\x64"
rename Rblas.dll Rblas.dll.orig
rename Rlapack.dll Rlapack.dll.orig
cd "C:\Program Files (x86)\IntelSWTools\compilers_and_libraries\windows\redist\intel64_win\mkl"
copy mkl_rt.dll "C:\Program Files\R\R-4.0.0\bin\x64\Rblas.dll"
copy mkl_rt.dll "C:\Program Files\R\R-4.0.0\bin\x64\Rlapack.dll"
copy mkl_intel_thread.dll "C:\Program Files\R\R-4.0.0\bin\x64"

making this known the PATH.

NLopt

Available from https://nlopt.readthedocs.io/en/latest/ with R counterpart from https://cran.r-project.org/web/packages/nloptr/index.html.

GNU Octave

It is available with,

sudo apt install octave

PSPP

Under Ubuntu, this can be made available with

sudo apt install pspp

For Fedora, we have

sudo dnf install pspp

which will install libpq, gsl, gtksourceview3, spread-sheet-widget as well, see https://apps.fedoraproject.org/packages/pspp.

Two simple SPSS command files example.sps and plot.sps can be called with

pspp example.sps 
psppire plot.sps

showing CLI and GUI, respectively. Related utilities are pspp-convert.

It is possible to compile it directly by using

  • gtksourceview 4.0.3 (4.4.0 is more demanding with Python 3.5, meson, Vala, etc.) and use PKG_CONFIG_PATH when appropriate
  • spread-sheet-widget-0.3
  • fribidi-1.0.8
  • GTKSOURVIEW_CFLAGS and GTKSOURVIEW_LIBS in the configuration.
export PREFIX=/rds/user/$USER/hpc-work
export GTKSOURCEVIEW_CFLAGS=-I${PREFIX}/includegtksourceview-4
export GTKSOURCEVIEW_LIBS="-L${PREFIX}/lib -lgtksourceview-4"
./configure --prefix=${PREFIX}
make
make install

note that it is necessary to comment on the statement kludge = gtk_source_view_get_type (); from src/ui/gui/widgets.c and to remove the PREFIX= speficiation in the Perl part of compiling, i.e,

cd perl-module
/usr/bin/perl Makefile.PL PREFIX=/rds/user/$USER/hpc-work OPTIMIZE="-g -O2 -I/rds-d4/user/$USER/hpc-work/include/fribidi -I/usr/include/cairo -I/usr/include/glib-2.0 -I/usr/lib64/glib-2.0/include -I/usr/include/pixman-1 -I/usr/include/freetype2 -I/usr/include/libpng15 -I/usr/include/uuid -I/usr/include/libdrm -I/usr/include/pango-1.0 -I/usr/include/harfbuzz  "

A more recent description is here, https://cambridge-ceu.github.io/csd3/applications/pspp.html.

python

A useful resource is code from Pattern Recognition and Machine Learning.

It is possible to conduct survival analysis with lifelines,

pip install lifelines

R

Fedora 31

The R-release, including both the compiled and source package, is built as follows,

sudo dnf install R
sudo dnf install R-devel

while the following are necessary to build the development version,

sudo dnf install gcc-c++
sudo dnf install gcc-gfortran
sudo dnf install pcre-devel
sudo dnf install java-1.8.0-openjdk-devel
sudo dnf install readline-devel
sudo dnf install libcurl-devel
sudo dnf install libX11-devel
sudo dnf install libXt-devel
sudo dnf install bzip2-devel
sudo dnf install xz-devel
sudo dnf install pandoc
sudo dnf install qpdf
sudo dnf install texlive-collection-latex
sudo dnf install texlive-collection-fontsextra
sudo dnf install texinfo-tex
sudo dnf install texlive-collection-fontsrecommended
sudo dnf install texlive-collection-latexrecommended
./configure

This is necessary since gcc 9 is available and required for CRAN package submission, e.g.,

# R-release to build
R CMD build gap
# R-devel to check
ln -s $HOME/R/R-devel/bin/R $HOME/bin/R-devel
R-devel CMD check --as-cran gap_1.1-22.tar.gz

For R-devel, these can be used explicitly,

export CC="/usr/bin/gcc"
export CXX="/usr/bin/g++"
export FC="/usr/bin/gfortran"
export CFLAGS="-g -O2 -Wall -pedantic -mtune=native"
export FFLAGS="-g -O2 -mtune=native -Wall -pedantic"
export CXXFLAGS="-g -O2 -Wall -pedantic -mtune=native -Wno-ignored-attributes -Wno-deprecated-declarations -Wno-parentheses"
export LDFLAGS="-L/usr/lib64"
R-devel CMD INSTALL gap_1.2.tar.gz

with check on foreign language calls or

R-devel CMD INSTALL --configure-args="
 CC=\"/usr/bin/gcc\" \
 CXX=\"/usr/bin/g++\" \
 FC=\"/usr/bin/gfortran\" \
 CFLAGS=\"-g -O2 -Wall -pedantic -mtune=native\" \
 FFLAGS=\"-g -O2 -mtune=native -Wall -pedantic\" \
 CXXFLAGS=\"-I/usr/include -g -O2 -Wall -pedantic -mtune=native -Wno-ignored-attributes -Wno-deprecated-declarations -Wno-parentheses\" \
 LDFLAGS=\"-L/usr/lib64\" gap_1.1-26.tar.gz
 ```
which is more restrictive than the default --as-cran above. A simpler setup is also possible with `~/.R/Makevars`, e.g.,
```bash
CC = gcc
CXX = g++
CXX11 = g++
FC = gfortran
F77 = gfortran
F90 = gfortran
CFLAGS = -std=c99 -I/usr/include -g -O2 -Wall -pedantic -mtune=native -Wno-ignored-attributes -Wno-deprecated-declarations -Wno-parentheses -Wimplicit-function-declaration
CXXFLAGS = -std=c++11

Another example is as follows,

module load texlive
./configure --prefix=/rds-d4/user/jhz22/hpc-work \
            --enable-R-shlib \
            CPPFLAGS=-I/rds-d4/user/jhz22/hpc-work/include \
            LDFLAGS=-L/rds-d4/user/jhz22/hpc-work/lib

On Fedora 35, we see the following messages from R CMD check gap_1.2.3-6.tar.gz,

Error(s) in re-building vignettes:
  ...
--- re-building ‘gap.Rmd’ using rmarkdown
Quitting from lines 273-279 (gap.Rmd)
Error: processing vignette 'gap.Rmd' failed with diagnostics:
X11 font -adobe-helvetica-%s-%s-*-*-%d-*-*-*-*-*-*-*, face 1 at size 5 could not be loaded
--- failed re-building ‘gap.Rmd’
--- re-building ‘shinygap.Rmd’ using rmarkdown
--- finished re-building ‘shinygap.Rmd’
--- re-building ‘jss.Rnw’ using Sweave
--- finished re-building ‘jss.Rnw’

SUMMARY: processing the following file failed:

  ‘gap.Rmd’

Error: Vignette re-building failed.
Execution halted

* checking PDF version of manual ... OK
* checking HTML version of manual ... NOTE
Skipping checking HTML validation: no command 'tidy' found
Skipping checking math rendering: package 'V8' unavailable
* checking for non-standard things in the check directory ... OK
* checking for detritus in the temp directory ... OK
* DONE

This is resolved by

sudo dnf install v8-devel
sudo dnf install xorg-x11-fonts*
Rscript -e 'install.packages(c("shniy","V8"),repos="https://cran.r-project.org")'

Ubuntu 18.04

The R environment is furnished with

sudo apt install r-base-core
sudo apt install r-base-dev

and R_LIBS is set from .bashrc

export R_LIBS=/usr/local/lib/R/site-library/

Note that in fact html.start() in R points to /usr/local/lib/R/library/ instead, see below example in MendelianRandomization.

To enable R-devel/package building, these are necessary

sudo apt install g++
sudo apt install gfortran
sudo apt install texlive
sudo apt install texlive-fonts-extra
sudo apt install texinfo
sudo apt install texlive-fonts-recommended
sudo apt install libreadline-dev

To set up bzip2, lzma/pcre, curl and then R assuming lapack is already installed,

# compile shared library Makefile-libbz2_so and then add -FPIC to CC and recompile
# bzip2
# make
# make install PREFIX=$SHOME

# xz
# ./configure --prefix=SHOME/xz-5.2.3
# make -j3
# make install

# pcre
# ./configure  --prefix=$SHOME  --enable-utf8

# curl
# ./configure  --prefix=$SHOME --with-ssl
# make && make install
./configure --prefix=/scratch/jhz22 --enable-R-shlib CPPFLAGS="-I/scratch/jhz22/include" LDFLAGS="-L/scratch/jhz22/lib"

Windows

To build packages on Windows, download Rtools from https://cran.r-project.org/ and install to C:\Rtools

rem 22/8/2019 JHZ

set path=C:\Program Files\R\R-3.6.1\bin;c:\Rtools\bin;%PATH%;c:\Rtools\mingw_64\bin;c:\Rtools\mingw_32\bin
set lib=c:\Rtools\mingw_64\lib;c:\Rtools\mingw_32\include
set include=c:\Rtools\mingw_64\include;c:\Rtools\mingw_32\include

We can then run R CMD INSTALL --binary gap, say.

It seems the --arch x84 option is very useful for using all available RAM; to make sure use call such as D:\Program Files\R\R-3.5.0\bin\x64\R.exe".

When this fails, remove large objects in your code and start R with --vanilla option.

To upgrade R, it is useful to install installr for its updateR().

Package installation

CRAN. It is typically done with install.packages()

install.packages("ggplot2",INSTALL_opts="--library=/usr/local/lib/R/site-library/")

Bioconductor. This is done with biocLite.

source("https://bioconductor.org/biocLite.R")
biocLite("packagename")

From R 3.5 or greater there is BiocManager,

if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
BiocManager::install()

See https://bioconductor.org/install/.

Lastly, it is possible with devtools::install_bioc().

GitHub. We could set this up via sudo apt install r-cran-devtools. This is then through devtools::install_github().

library(devtools)
install_github("MRCIEU/TwoSampleMR",args="--library=/usr/local/lib/R/site-library",force=TRUE)

with dedicated location(s); however this is not always the case and an alternative is to use

sudo R CMD INSTALL <packed/unpacked package> -l $R_LIBS

to install into $R_LIBS.

It is possible to point to a package, locally or remotely, e.g,

install.packages("http://cnsgenomics.com/software/gsmr/static/gsmr_1.0.6.tar.gz",repos=NULL,type="source")

whose first argument is a URL.

Multiple precision arithmetic. This is modified from notes on SCALLOP-INF analysis.

sudo apt install libmpfr-dev
R --no-save <<END
require("Rmpfr")
z <- 20000
Rmpfr::pnorm(-mpfr(z,100), mean = 0, sd = 1, lower.tail = TRUE, log.p = FALSE)
Rmpfr::pnorm(-mpfr(z,100), mean = 0, sd = 1, lower.tail = TRUE, log.p = TRUE)
# competitor
-log10(2)-pnorm(-z,lower.tail=TRUE,log.p=TRUE)/log(10)
END

The function pnorm(-abs(z), lower.tail=TRUE, log.p=TRUE) in R is doing surprisingly well and now an R/gap utility function.

tidy.R. The following code formats R source codes according to the R session,

function tidy()
{
  export input=$1
R --vanilla <<END
  options(keep.source = FALSE)
  input <- Sys.getenv("input")
  source(input)
  dump(ls(all = TRUE), file = paste0(input,"_out"))
END
}

Package reinstallation

For instance to replace packages under gcc 4.4.7 to gcc 5.4.0, one can resinstall all packages as in /scratch/jhz22/R to /home/jhz22/R,

R --no-save <<END
from <- "/scratch/jhz22/R"
to <- "/home/jhz22/R"
pkgs <- unname(installed.packages(lib.loc = from)[, "Package"])
pkgs
install.packages(pkgs, lib=to, repos="https://cran.r-project.org")
END

Mirror

We can mirror R packages from $HOME to /scratch as follows,

rsync -avrzP /home/$USER/R /scratch/$USER/R

Reinstallation

This could be done with

update.packages(checkBuilt = TRUE, ask = FALSE) 

Specification of repository

One may see error message on package installation

> install.packages("plotly")

--- Please select a CRAN mirror for use in this session ---

Error in structure(.External(.C_dotTclObjv, objv), class = "tclObj") : 

  [tcl] bad pad value "2m": must be positive screen distance.

but can be avoided with specificatino of repository.

> install.packages("plotly", repos="https://cran.r-project.org")

RStudio

The distribution has problem loading or creating R script, so it is tempting to install from https://github.com/rstudio/rstudio/. This involves running scripts under directory dependencies/,

./install-dependencies-debian --exclude-qt-sdk

and then the following steps,

mkdir build
cd build
cmake .. -DRSTUDIO_TARGET=Desktop -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=/usr/local/lib/rstudio

However, there is error with Java and Java 8 is required, see https://tecadmin.net/install-oracle-java-8-ubuntu-via-ppa/.

sudo add-apt-repository ppa:webupd8team/java
sudo apt-get update
sudo apt-get install oracle-java8-installer
sudo apt-get install oracle-java8-set-default
java -version

However, compile error is still persistent except when dropping the option --exclude-qt-sdk but unloadable.

It is therefore recommended to get around with RStudio daily builds, https://dailies.rstudio.com/.

SageMath

sudo apt install sagemath

stan

cmdstan is now available from https://github.com/stan-dev/cmdstan along with other repositories there. Interfaces are listed at http://mc-stan.org/users/interfaces/index.html.

Information on installing RStan is described here,

https://github.com/stan-dev/rstan/wiki/Installing-RStan-on-Linux

On our HPC system under gcc 4.8.5 there are error message

> library(rstan)
Loading required package: ggplot2
Registered S3 methods overwritten by 'ggplot2':
  method         from
  [.quosures     rlang
  c.quosures     rlang
  print.quosures rlang
Loading required package: StanHeaders
Error: package or namespace load failed for ‘rstan’ in dyn.load(file, DLLpath = DLLpath, ...):
 unable to load shared object '/rds-d4/user/jhz22/hpc-work/R/rstan/libs/rstan.so':
  /usr/lib64/libstdc++.so.6: version `GLIBCXX_3.4.20' not found (required by /rds-d4/user/jhz22/hpc-work/R/rstan/libs/rstan.so)
> q()

which can be resolved with module load gcc/5.2.0 before invoking R.

For error message C++14 standard requested but CXX14 is not defined we modify $HOME/.R/Makevars as follows,

CXX14 = g++ -std=c++1y -fPIC

see https://github.com/stan-dev/rstan/issues/569 but adding -fPIC and as in unixOBD below.

unixODBC

It is quite standard to install, i.e.,

wget ftp://ftp.unixodbc.org/pub/unixODBC/unixODBC-2.3.7.tar.gz
tar xvfz unixODBC-2.3.7.tar.gz
cd unixODBC-2.3.7
./configure --prefix=/scratch/jhz22
make
make install

There have been many discussions regarding "C++11 standard requested but CXX11 is not defined" and this could be fixed with changes to $R_HOME/etc/Makeconf such that

CXX11 = g++ -std=c++11 -fPIC

then

module load gcc/5.2.0
R CMD INSTALL odbc

This is necessary for gtx for instance.

zlib

Try

sudo apt-get install libz-dev