[Scipy-tickets] [SciPy] #1746: scipy.optimize -> ImportError: No module named _components

SciPy Trac scipy-tickets@scipy....
Sun Oct 14 17:45:13 CDT 2012


#1746: scipy.optimize -> ImportError: No module named _components
------------------------------------------------------------+---------------
 Reporter:  goneasync                                       |       Owner:  somebody   
     Type:  defect                                          |      Status:  needs_info 
 Priority:  normal                                          |   Milestone:  Unscheduled
Component:  scipy.optimize                                  |     Version:  0.10.1     
 Keywords:  _components cs_graph_components csgraph sparse  |  
------------------------------------------------------------+---------------

Comment(by goneasync):

 Hello, hopefully my naivete isn't wasting too much of your time, but many
 thanks for the support. Hopefully by build log you meant the raw build
 output via standard redirect (attached) -- I couldn't find any log files.

 I built with what I thought was the standard `sudo python3.2 setup.py
 install` from within the un-packed .tar.gz top level directory, barring my
 finding any exceptions of suggestions in the .txt files. I *think* per my
 experience in trying to get several different things installed, that these
 days 2to3 seems to be run implicitly when trying to build like this; this
 is after trying and failing to follow what appears to be dated advice on
 using 2to3 more overstly to get 2.X-optimised third party packages
 installed on my 3.2 system.

 Could this be something to do with 10.6 vs 10.7 incompatibility, i.e.
 there's something about scipy0.10.1 that hard-codes 10.6? I couldn't find
 any advice on setting config parameters or environment variables to
 specify 10.7, but as I mentioned I was even less successful with getting
 scipy0.11.0 to install.

 Many thanks once again!

-- 
Ticket URL: <http://projects.scipy.org/scipy/ticket/1746#comment:3>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.


More information about the Scipy-tickets mailing list