Re: Very slow reading of a 2D matirx

[Paul Kienzle]

[moved from help list]

On Oct 22, 2005, at 7:38 PM, John W. Eaton wrote:

> On 22-Oct-2005, Paul Kienzle wrote:
>
> | On Oct 21, 2005, at 8:37 AM, John W. Eaton wrote:
> | > Should Octave export a variable that holds the full name of the
> | > current script (or even function) file while it is executing?
> |
> | That would be useful in some circumstances, but I wouldn't want every
> | function invocation to be slower just to support it.
>
> I think we already have the variable internally for other purposes.
> So if we implement access to it as a function instead of modifying the
> value of a variable in the symbol table, I don't think there would be
> any penalty unless you request the value.

The usual call when you are looking for the path is 'which fun'.
Shouldn't we extend this do the correct thing from within 'fun'?

I was playing with which and private functions.  I defined
two files.

testing.m:

function testing,
  private_max
  which private_max
  which testing

function private_max, 'testing private_max'

private_max.m:

function private_max, 'private_max'


From the command line I got the following result:

> octave:1> testing
> ans = private_max
> private_max is the user-defined function from the file
> /Users/paulkienzle/junk/private_max.m
> testing is the user-defined function from the file
> /Users/paulkienzle/junk/testing.m


Shouldn't this have printed 'testing private_max' and
reported that private_max is defined within testing.m?

I'm running Mac OS 10.3 octave 2.1.69 from hpc.sf.net.

Thanks,

        - Paul