[Starcluster] NumPy and SciPy tests fail on Ubuntu 9.10 Karmic

Justin Riley jtriley at MIT.EDU
Mon Apr 19 14:43:28 EDT 2010


Hi Andrew,

> Justin, you could use stdeb to custom install Python packages (e.g.
> numpy/scipy) while letting dpkg know about it -- thus preventing
> clobbering between dpkg and distutils. See
> http://github.com/astraw/stdeb . Hopefully it's pretty straightforward
> to use.

Ahhh that's very handy. Similar to gpypi in Gentoo. Thanks for the link and 
for creating it :D

I do have dpkgs for the custom openmpi installation, however, apt-get still 
wants to override my custom packages installed via "dpkg -i" when doing apt-
get update. Do you know how to fix that? Seems like something trivial I'm 
missing. 

I'll give stdeb a shot and hopefully create some custom numpy/scipy debs for 
the ami. However, the harder package to customize I suspect will be lapack 
which is pretty much the entire reason for custom compiling numpy/scipy debs 
in the first place. I noticed you have or had a repo for ubuntu numpy/scipy 
(from http://www.scipy.org/Installing_SciPy/Linux). Did your packages use the 
default debian lapack?

Thanks again for informing me of stdeb!

~Justin



On Monday 19 April 2010 2:21:30 pm Andrew Straw wrote:
> Justin Riley wrote:
> > Hi Damian,
> > 
> > Good to hear you fixed the issue. I will certainly need to add a note
> > about installing numpy/scipy/lapack/openmpi from ubuntu. These packages
> > have been custom compiled and installed (see
> > http://starcluster.scripts.mit.edu/~starcluster/wiki/index.php?title=Star
> > Cluster_AMI_Cookbook). Openmpi happens to be recompiled into a deb
> > package so it's easy to reinstall, however, numpy/scipy/lapack require a
> > bit more work if you clobber them with apt-get.
> 
> Justin, you could use stdeb to custom install Python packages (e.g.
> numpy/scipy) while letting dpkg know about it -- thus preventing
> clobbering between dpkg and distutils. See
> http://github.com/astraw/stdeb . Hopefully it's pretty straightforward
> to use.
> 
> -Andrew



More information about the StarCluster mailing list