Date: Fri, 29 Mar 2024 11:45:01 -0400 (EDT) Message-ID: <1084324689.164.1711727102040@esjava53.uits.iu.edu> Subject: Exported From Confluence MIME-Version: 1.0 Content-Type: multipart/related; boundary="----=_Part_163_1983461340.1711727101995" ------=_Part_163_1983461340.1711727101995 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Content-Location: file:///C:/exported.html
We install various python modules for use on the Luddy unified Linux sys= tems. In addition, we have multiple versions of python available via = enviro= nment modules. The default version of python that is installed ma= y have fewer available modules than the versions available via modules so y= ou are encouraged to try the other versions using environment modules.= p>
If you need a python module installed, please just subm= it a service request letting us know what modules you need. In yo= ur request, please let us know if you are using the stock version of python= or a newer version we make available via environment modules. The ad= vantages of having the IT staff install python modules centrally include 1)= you don't have to do it yourself and take up your own disk quota, 2) they = are available to all users, and 3) they are available on all of the unified= linux systems we manage.
In some cases, you may want to just install python modules yourself.&nbs= p; If you want to do this, it is very simple and does not require any speci= al root/sudo permissions. All you need to do is just use the --user o= ption to pip and it will be installed in your own home directory and be aut= omatically accessible by you (but not other users). For example, if y= ou wanted to install a module into your own python library directory you wo= uld just run:
pip ins= tall --user package_name
You just need to replace package_name with the official name of= the python module.
When you use the --user option, pip will install the module into a defau= lt location for that system. For example, on the RHEL 7 systems with = the stock python 2.7, the module will end up being written into your ~/.loc= al/lib/python2.7 and/or ~/.local/bin directories.
You can also use python virtualenv to set up a local python environment = where you can install your own modules. UITS has a good KB page avail= able: Use Virtualenv to customize Python
The instructions in that KB page are for the UITS supercomputers (eg. Bi= g Red, etc) so it isn't exactly the same as on the Luddy systems but is nea= rly identical but with these differences:
With these minor exceptions, the instructions should be the same. = Note that you should not use the --user argument to pip when using a virtua= lenv, as noted in that UITS KB page.
You an also use conda to set up a local python environment. UITS h= as a good KB page available: Install packages in a conda environment on = IU's high performance computers
The instructions in that KB page are for the UITS supercomputers (eg. Bi= g Red, etc) so it isn't exactly the same as on the Luddy systems but is nea= rly identical. The differences of note on our systems are:
With those exceptions, the instructions should be very similar on the Lu= ddy systems.