[SciPy-dev] Using IPython

Janko Hauser jhauser at ifm.uni-kiel.de
Mon Feb 18 03:08:23 CST 2002


eric writes:
 > 
 > I'd like to say yes -- and it should be possible.  However, there are
 > implications for the testing framework.  In the past when we deleted misc in
 > __init__.py after all its stuff was imported, the testing frame work failed to
 > find the test suites for misc -- and died a premature death.  This probably can
 > be addressed, but not sure how at the moment.
 > 
Ok will look into this.

 > > At the moment there are 300 definitions in scipy.
 > 
 > There does need to be some more thought put into what is put in the top level.
 > I think it is necessarily a large number of functions, but we could certainly
 > trim some fat if it is found.
 > 
 > I've often wished that there was a "magic variable" in modules that one could
 > define to omit names from the name-completion tools in GUI's.  

There is the __all__ directive in recent pythons, which allows you to
specify, which names should be exported from a module. Haven't tested
this at the command prompt, will also do this and report back.

__Janko



More information about the Scipy-dev mailing list