I agree that the structures themselves shouldn't be part of the math
library. But having plot and the functions that calculate statistics
accept objects that implement generic interfaces is a good idea.
Neil ⊥
On 06/30/2012 07:50 PM, Ray Racine wrote:
Not aligned with the original request I know ... but I'd like to see
some a foundation for a Racket "data" library along the lines of Pandas.
And then then an additional layer such that the statistics and plot
libraries just works with it.
http://pandas.pydata.org/
On Sat, Jun 30, 2012 at 8:31 PM, Neil Toronto <neil.toro...@gmail.com
<mailto:neil.toro...@gmail.com>> wrote:
Being a glutton for punishment, I've decided to write a `math'
collection to be shipped with Racket. I'm writing it in Typed
Racket, so Typed Racket programs that `(require math)' can apply
mathematical functions with no overhead and regular Racket programs
will incur only the usual contract checks. Also, Vincent's TR
optimizer will fill the math collection with rainbows and sunshine.
I know what *I* want in a math library: statistics (for my research)
and basic linear algebra (for `plot'). What do *you* want?
(FWIW, this started with Antonio asking for inverse hyperbolic
functions on the dev mailing list. I just did those today, and
factorial, log-factorial, gamma, and log-gamma yesterday.)
Neil ⊥
____________________
Racket Users list:
http://lists.racket-lang.org/__users
<http://lists.racket-lang.org/users>
____________________
Racket Users list:
http://lists.racket-lang.org/users
