software:python-canopy:python-canopy

This is an old revision of the document!


Canopy Express provides an open, intuitive environment for scientific and analytic computing. See acomplete list of packages available in Canopy Express. Most of the packages not included in Canopy express are based on the most current version, old versions can be loaded if needed.

Please review the Enthought Academic License Agreement before using Canopy Express.

"Permitted Academic Use" shall mean use of the Software for internal academic teaching or research purposes, and explicitly excludes any use of the Software for commercial purposes including commercial research purposes.

To start using Canopy Express, use VALET

$ vpkg_require canopy
Performing first-time setup of Canopy...SUCCEEDED
Adding package `canopy/1.3.0` to your environment

The first time you use Canopy, it will install a virtual core development environment of Canopy Express in your home directory. It will take approximately 18M of space. It will add the appropriate directories to your PATH.

You can use the Canopy Express command line interface (CLI) to do your development on the head node, however it is always a good practice to run longer jobs using qsub for batch jobs. For example, on Mills for account traine, it would like this to use Canopy Express ipython

[traine@mills ~]$ vpkg_require canopy
Adding package `canopy/1.3.0` to your environment
[traine@mills ~]$ echo $PATH
/home/1200/.canopy-1.3.0/Canopy_64bit/User/bin:/opt/shared/canopy/1.3.0/_bin:/usr/lib64/qt-3.3/bin:/opt/bin:/home/software/OpenGridScheduler/2011.11p1/bin/linux-x64:/usr/local/bin:/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/sbin:/home/1200/bin
[traine@mills ~]$ which ipython
~/.canopy-1.3.0/Canopy_64bit/User/bin/ipython
[traini@mills ~]$ ipython
Python 2.7.6 | 64-bit | (default, Jan 29 2014, 17:35:36)
Type "copyright", "credits" or "license" for more information.

IPython 1.1.0 -- An enhanced Interactive Python.
?         -> Introduction and overview of IPython's features.
%quickref -> Quick reference.
help      -> Python's own help system.
object?   -> Details about 'object', use 'object??' for extra details.

In [1]:
If you are producing graphics files such as .png and even if you are not displaying them using show, you still need to make sure X11 is available by using vpkg_require x11, otherwise you will get an error for the graphics library. For example,
import matplotlib._png as _png
ImportError: libpng12.so.0: cannot open shared object file: No such file or directory
Sample commands here demonstrate example for : (cluster mills.hpc, user traine in workgroup it_css). However, same commands are also applicable for cluster farber

If you need to use the GUI (Graphical User Interface) to use the interactive editor or package manager to install a package not currently available in the virtual core development environment of Canopy Express, then you must do so on a compute node by using qlogin and make sure you are running X-Windows. For example, on Mills for account traine in workgroup it_css, it would look like this

[traine@mills ~]$ workgroup -g it_css
[(it_css:traine)@mills ~]$ qlogin
Your job 485034 ("QLOGIN") has been submitted
waiting for interactive job to be scheduled ...
Your interactive job 485034 has been successfully scheduled.
Establishing /opt/shared/OpenGridScheduler/local/qlogin_ssh session to host n017 ...
[traine@n017 ~]$ vpkg_require canopy
Adding package `canopy/1.3.0` to your environment
[traine@n017 ~]$ vpkg_require x11
Adding package `x11/RHEL6.1` to your environment
[traine@n017 ~]$ canopy

A Welcome to Canopy window should be displayed. If you try to run the Canopy Express GUI on the head node, it will crash with a segmentation fault.

In the package manager, you may see packages which indicate CANOPY and not FREE, when you click on these packages listed at CANOPY you will not have an Install option, only Subscribe. However if you click on More info, it will list dependencies and a Show button at the bottom of the window to show other versions available that are free to install. This is typical to allow older versions of a package to be installed, just not the latest version such as Mayavi.

Please avoid installing updates you may see when using the package manager as this will install a full version of the package and dependencies in your home directory thus potentially putting your account over quota. Even a small number of updates (< 10) caused the 18M base virtual core development environment to grow to 600M. Instead of installing updated, please send an email to consult@udel.edu and indicate which cluster to request the virtual core development environment to be updated.

If you are loading Canopy express to build a package based on Enthought's Python Distribution, then you may need to load other libraries and/or packages. For example, if you were trying to build the package for openmpi and gcc, your VALET setup might look like this

$ vpkg_devrequire canopy
Adding package `canopy/1.3.0` to your environment
$ vpkg_devrequire gcc/4.6
Adding package `gcc/4.6.2` to your environment
$ vpkg_devrequire openmpi/1.4.4-gcc
Adding package `openmpi/1.4.4-gcc` to your environment

You may want to reset your Canopy Express, perhaps to get an updated virtual environment or a newer version to minimize the disk usage in your home directory. You can do this by removing Canopy Express setup in your home directory. For example, in traine's home directory for Canopy Express version 1.3, it would look like this

[traine@mills ~]$ rm -r .canopy
[traine@mills ~]$ rm -r .canopy-1.3.0
[traine@mills ~]$ rm -r canopy

Once the Canopy Express setup directories in your home directory are removed, the next time you use vpkg_require canopy it will install the latest virtual core development environment of Canopy Express in your home directory.

  • software/python-canopy/python-canopy.1524761571.txt.gz
  • Last modified: 2018-04-26 12:52
  • by sraskar