..Bmax and scientific use...

BlitzMax Forums/BlitzMax Programming/..Bmax and scientific use...

Naughty Alien

(Posted 2009) [#1]

..if you would like to use Bmax in such way, is there any serious limitation in Bmax for such use??

Tommo

(Posted 2009) [#2]

I'm sure you will need a arbitrary precision math module.
Check brucey's BaH.MAPM

Brucey

(Posted 2009) [#3]

Alas, BlitzMax is a "Games Programming Language" :-(

Warpy

(Posted 2009) [#4]

Yeah, I don't think it would be a good fit - you'll need arbitrary precision maths and there are no libraries of useful scientific functions.

Have you seen SciPy?

r	(Posted 2009) [#5]

try this test code :

Local value:Double
Local i,j,k:Int

For i=1 To 10
For j=1 To 1000
For k=1 To 1000
value=Tan(ATan(10^(Log10(Sqr(value*value)))))+1
Next
Next
Next

Print String(value)

the result would be : 10000001,0001 in fortran
blitzmax : 100009576,2290
delphi : 10000001,0011137
purebasic : 1708221.5000000533
gfabasic 32 : 10000001,0010595
freebasic : 10000000,92740968

beanage

(Posted 2009) [#6]

poorbasic^^ .. well, i think it highly depends on what u mean with "scientific". I'd hardly use it for sattelite or particle accelerator control, but testing NNs works great!

Brucey

(Posted 2009) [#7]

BaH.MAPM is good for arbitrary precision maths.
BaH.MathToolkit has a large assortment of statistical distributions, including Binomial, Cauchy, Pareto, Weibull and Students.
As well as a selection of "special functions", including Bessel, Neumann, Gamma, Zeta and Factorial.
BaH.muParser is an extendible math parser.

Haven't got around to doing any others yet - finding non-GPL'd libraries is difficult.

All available from HERE.

(Note, that mathtoolkit also requires BaH.Boost - there's a combi download available with both)

srvaldez

(Posted 2009) [#8]

with the variable value not initialized and therefore assumed to be 0, why does it work?
shouldn't log10(0) return a NaN?

Azathoth

(Posted 2009) [#9]

Floyd

(Posted 2009) [#10]

shouldn't log10(0) return a NaN?

It actually returns -Infinity, the limiting value of Log10( x ) as x goes to 0 through positive numbers.

Similarly, 10^( -Infinity ) is 1/Infinity, i.e. 0.