[TIB] Re: Integral Variables...


[Prev][Next][Index][Thread]

[TIB] Re: Integral Variables...



Agh.  Sorry.  Travis is a much more common name than Tavis is...just didn't
bother to read it carefully.  Sorry again.

So it seems that, while exact(int(x)) does force it to be an integer, it
does not set the data type of the variable to be an integer, so the
calculations have no speed difference.  It does, however, restrict the
answers to set J.  So...it does have its uses.  Just not for what I was
hoping for.  So...now that I can make an integer variable...anyone have some
truly special sort of integer operand functions?

I thought as much.  Ah, well.  I'll find plenty of other ways to play with
it, I'm sure.

And, just for the record, trying to leaf your way through a 8x20 matrix
is -so- much faster than a 24x60.  Woo subMat().  Again.

----- Original Message -----
From: "Tavis Segura" <tavis@hotmail.com>
To: <ti-basic@lists.ticalc.org>
Sent: Wednesday, April 09, 2003 12:16 PM
Subject: [TIB] Re: Integral Variables...


>
>
> >
> >To [Tavis]: I'm only using the basic arithmatic operators,
> >so I don't really have a worry about wandering outside J.
>
> It is true, but I'm sure anyone reading the archives might
> have other ideas whenever they read these, so that point
> needed to be addressed anyway.
>
> Going slightly off topic, I want to remind people my name is
> spelled Tavis, and is pronounced with a long "a", as in "tape".
> Spelling may not seem important when talking directly, but it
> can be a big deal if anyone tries to contact me outside of
> this list.
>
> >Ah, well.  I'll let everyone know the speed increase (even if it is only
2%
> >or something) when I get back.  Keep a record of all this when the next
kid
> >wants to use it.
>
> Sure thing! Maybe you should time the instructions for int, exact,
> and exact(int()) while you're at it. It might help people determine
> whether their algorithms will perform better with approximate versus
> exact answers.
>
> On a side note about performance, I recently modified some
> functions in Aim to use less precision in order to gain speed,
> such as simplifying expressions containing pi to numeric values
> when very few significant digits are needed. Such was done in
> the terrain generator, where only whole number accuracy affected
> the visible height of the terrain.
>
> >On that note, use subMat()!  What a wonderful function when
> >dealing with data in matrices.
> >
>
> I just might try that sometime soon... I need to reorganize my
> data before I can take advantage of that function, though.
> Tavis
>
> _________________________________________________________________
> Add photos to your e-mail with MSN 8. Get 2 months FREE*.
> http://join.msn.com/?page=features/featuredemail
>
>




References: