«

Hi,

I have a problem with calculating expression, in which I have to use

BigDecimal1 ^ BigDecimal2


BigDecimal.pow() isn't suitable - as far as I know, it requires int as
an parameter, and both my numbers are BigDecimals.


Does anyone know how to handle it?

sfetr

wrote:
> Hi,
>
> I have a problem with calculating expression, in which I have to use
>
> BigDecimal1 ^ BigDecimal2
>
>
> BigDecimal.pow() isn't suitable - as far as I know, it requires int as
> an parameter, and both my numbers are BigDecimals.
>
>
> Does anyone know how to handle it?

bd1.pow(bd2.intValue()) looks like a good starting point.
If bd2 has a fractional part, you'll need to work harder --
perhaps Math.pow(bd1.doubleValue(), bd2.doubleValue()) would
be of use, followed by conversion from double to BigDecimal.

Seems like a strange thing to want to do, though. I hope
you've got lots of memory ...

--
Eric Sosman
lid

Eric Sosman wrote:
> wrote:
>> Hi,
>>
>> I have a problem with calculating expression, in which I have to use
>>
>> BigDecimal1 ^ BigDecimal2
>>
>>
>> BigDecimal.pow() isn't suitable - as far as I know, it requires int as
>> an parameter, and both my numbers are BigDecimals.
>>
>>
>> Does anyone know how to handle it?
>
> bd1.pow(bd2.intValue()) looks like a good starting point.
> If bd2 has a fractional part, you'll need to work harder --
> perhaps Math.pow(bd1.doubleValue(), bd2.doubleValue()) would
> be of use, followed by conversion from double to BigDecimal.
>
> Seems like a strange thing to want to do, though. I hope
> you've got lots of memory ...
>

Even if the double approach does not give enough accuracy, it might be a
good starting point for an iterative improvement algorithm such as
Newton-Raphson.

However, it definitely needs to be dome carefully, with attention to
trade-offs between accuracy and memory.

Patricia

Thanks a lot, problem vanished