Re: DecimalFormat

["James H. H. Lampert" <jamesl@xxxxxxxxxxx>]

"Xu, Weining" wrote:

> May be I didn't say clearly on my question.  Add 0.005 to the number prior
> to the rounding would cause the number always add one to the last decimal,
> such as 2.3237 became 2.33, which is not what I want.  I just need to do
> normal rounding, meaning if 5 and up goes up, 4 and below takes off.

Yes, if you added 0.005 before applying a TRUE ROUNDING to the number,
it would behave as specified, as true rounding is defined as adding (in
this case) 0.005, then truncating (which always rounds down), and you
would effectively be adding 0.005 twice, then truncating. Thus, if the
number is being truncated, and you want true rounding, add the 0.005; if
the number is being truly rounded and you want truncation, you could
subtract 0.005.

As to whether DOUBLE would work better, it might. There are certain
numbers that have an exact decimal representation, but simply cannot be
represented exactly in floating-point binary (just as there are some
numbers, e.g., 1/3, that can be exactly represented as indicated
quotients, but cannot be exactly represented in decimals).

--
James H. H. Lampert
Professional Dilettante
http://www.hb.quik.com/jamesl
http://members.hostedscripts.com/antispam.html
http://www.thehungersite.com

Read My Lips: No More Atrocities!