Re: Most significant digit number formatting

If it is a numeric data column -- you probably want to use the "round"
function:

SELECT round(1200.01, 3);
SELECT round(12.009, 2);

"Berend Tober" <btober(at)seaworthysys(dot)com> wrote in message
news:4379F997(dot)6050703(at)seaworthysys(dot)com(dot)(dot)(dot)
> Say I want to format calculated numeric output to uniformly have a
> specific number of decimal places, say 3 digits right of the decimal
> point. My current understanding is that the TO_CHAR function is the
> appropriate way to accomplish this kind of numeric formatting. So for
> instance I might write
>
> SELECT
> project_number,
> TO_CHAR(SUM(labor_hours), '999.999') AS total_hours
> FROM labor_data
>
> This is great as far as accomplishing the decimal part formatting, but
> it is only good provided I've included enough place holders for the
> integer part, i.e., in this example for numbers less than 1000, e.g.,
>
> project_number total_hours
> 05-100 ###.### (actual value is 10810.5)
> 05-125 285.000
> 05-150 404.500
> 05-200 44.000
> 05-54 66.000
> 05-59 ###.### (actual value is 2245.75)
>
> So what I'm asking for is advice on how to create the least-significant
> digit formatting specifically, but without having to worry about
> exceeding the most-significant digit formatting specification. I don't
> see that explained in the documentation on TO_CHAR.
>
> I suppose on approach might to guess what the biggest number might be,
> and then include an order of magintude larger, e.g.
> TO_CHAR(SUM(labor_hours), '999999.999') . But you know, good old Dr.
> Murphy, will undoubtly intervene and inevitably create a situation in
> which whatever reasonable provisions were made originally, the limits
> will at some point be exceeded, causing the world as we know it to come
> to an end.
>
> Regards,
> Berend Tober
>
>
>

--------------------------------------------------------------------------------

>
> ---------------------------(end of broadcast)---------------------------
> TIP 6: explain analyze is your friend
>