Re: [Maxima-discuss] bfloat issue

[Dimiter P. <dim...@gm...>]

Hi Stavros,

I am trying to define infinitesimal values to use in symbolic calculation.
Obviously  numerically such values are evaluated to 0.


>> float(zeroa);
>>0

>> float(zeroa+5/3);
>> 1.666666666666666

>> bfloat(zeroa+5);

>> zeroa+5.0b0


If a mixed expression having an infintesimal and numeric part evaluates in
float it is OK but if for whatever reason big float precision is necessary
the approach will not work because of said limitation.

best regards,

Dimiter


On Fri, Sep 26, 2014 at 5:33 AM, Stavros Macrakis (Σταῦρος Μακράκης) <
mac...@al...> wrote:

> This is not surprising. numerval assumes that a numeric constant can be
> defined by a single floating-point value, and is explicitly documented to
> work with the 'numer' flag (which uses machine floating point). How would
> you expect it to work with bfloats? What if you're calculating with more
> precision than the numerval value?
>
> Are you sure numerval is the right mechanism for the calculation you're
> doing? What exactly are you trying to do?
>
>                   -s
>
> On Thu, Sep 25, 2014 at 6:02 PM, Dimiter Prodanov <dim...@gm...>
> wrote:
>
>> Hello,
>>
>> When I define
>> numerval (val,0)
>> for float (var) it works fine but bfloat does not recognize the
>> declaration.
>>
>> best regards,
>>
>> Dimiter
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
>> Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
>> Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
>> Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
>>
>> http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
>> _______________________________________________
>> Maxima-discuss mailing list
>> Max...@li...
>> https://lists.sourceforge.net/lists/listinfo/maxima-discuss
>>
>>
>