[erlang-questions] math:pow(X, Y).

[Hynek Vychodil]

I bet it is because HiPE will optimize it in different way and result
will be faster.

2012/3/20 Aleksandr Vinokurov <aleksandr.vin@REDACTED>:
> Richard, what the point of writing two equal funs? rtpi and rtpf? Really.
>
>
>
> 20.03.2012, в 5:35, "Richard O'Keefe" <ok@REDACTED> написал(а):
>
>>
>> On 20/03/2012, at 6:17 AM, Yves S. Garret wrote:
>>
>>> Hi all,
>>>
>>>   That's the standard method that I use to raise a value to a power.  I compared this to the way Python does raising to a power and this is the result that I get:
>>>
>>> Erlang:
>>> 11> math:pow(42, 909).
>>> ** exception error: bad argument in an arithmetic expression
>>>     in function  math:pow/2
>>>        called as math:pow(42,909)
>>>
>>> Python:
>>> http://bin.cakephp.org/view/1006418268
>>>
>>> Now.  Why is it in Erlang's case I get a float as a return value?
>>
>> Because you explicitly called something that is *defined* to return
>> floating point values.  Basically, the math: module consists of
>> wrappers around the C <math.h> library.  math:pow(1, 1) => 1.0
>> because math:pow/2 is something you call when you *want* a floating
>> point result.
>>
>>> Is there a method that only returns a long?
>>
>> What's a "long"?
>>
>> -module(rtp).
>> -export([rtp/2]).
>>
>> %% rtp(Base, Power) returns Base ** Power.
>> %% The answer has the same type as Base.
>> %% If Base is an integer, Power must be non-negative.
>>
>> -spec rtp(integer(), integer()) -> integer()
>>    ;    (float(),   integer()) -> float().
>>
>> rtp(Base, Power)
>>  when is_integer(Base), is_integer(Power), Power >= 0 ->
>>    rtpi(Base, Power, 1);
>> rtp(Base, Power)
>>  when is_float(Base), is_integer(Power), Power >= 0 ->
>>    rtpf(Base, Power, 1.0);
>> rtp(Base, Power)
>>  when is_float(Base), is_integer(Power), Power < 0 ->
>>    1.0 / rtpf(Base, -Power, 1.0).
>>
>> -spec rtpi(integer(), integer(), integer()) -> integer().
>>
>> rtpi(Base, Power, R) when Power > 2 ->
>>    rtpi(Base * Base, Power bsr 1,
>>        case Power band 1
>>          of 0 -> R
>>           ; 1 -> R*Base
>>        end);
>> rtpi(Base, 2, R) -> R * (Base * Base);
>> rtpi(Base, 1, R) -> R * Base;
>> rtpi(   _, 0, R) -> R.
>>
>> -spec rtpf(float(), integer(), float()) -> float().
>>
>> rtpf(Base, Power, R) when Power > 2 ->
>>    rtpf(Base * Base, Power bsr 1,
>>        case Power band 1
>>          of 0 -> R
>>           ; 1 -> R*Base
>>        end);
>> rtpf(Base, 2, R) -> R * (Base * Base);
>> rtpf(Base, 1, R) -> R * Base;
>> rtpf(   _, 0, R) -> R.
>>
>>
>> _______________________________________________
>> erlang-questions mailing list
>> erlang-questions@REDACTED
>> http://erlang.org/mailman/listinfo/erlang-questions
> _______________________________________________
> erlang-questions mailing list
> erlang-questions@REDACTED
> http://erlang.org/mailman/listinfo/erlang-questions



-- 
Hynek Vychodil
BI consultant

GoodData
náměstí 28. října 1104/17, 602 00, Brno - Černá Pole
Office:   +420 530 50 7704
E-mail:  hynek@REDACTED
Web:     www.gooddata.com