Starting with the academic year 2017/2018 all class workstations will be configured to offer maximum flexibility to python users. Below is a summary of the most important settings that you should take into account when working with python.
All workstations will run a pre-generated windows 10 image containing python, spyder, anaconda, etc..The operating system will be upgraded typically once a year and so will the python packages installed therein.
Because upgrading python packages once a year might not be sufficient for certain tasks, each workstation will source any python package present in a central repository before sourcing any locally installed package. This set up has the effect of overriding any locally installed python package should an upgrade be available in the central repository. Because the central repository can also host packages that are not locally installed at all, it can effectively act as an expansion of the local python library.
The central python repository is only writeable by the administrator and is located on the so-called T: drive
.
If your workstation cannot map the T: drive
, please contact the administrator or the helpdesk@physics.
To offer maximum flexibility and to allow users to do their programming as they wish, any python packages present in their home space in P:\AppData
will be sourced first to allow overriding any of the centrally or locally installed python packages. For instance if a user wanted to install a package, they could execute the following lines
# open anaconda prompt set APPDATA=P:\AppData pip install <whatever package> --user
Similarly, the upgrade of an already existing package can be done in this way
# open anaconda prompt set APPDATA=P:\AppData pip install -U <whatever package> --user
Add –no-deps
if you want to exclude all dependencies from the install process.
If the P: drive
is not mapped on your workstation, please contact the administrator or the helpdesk@physics.