"Phil Cook" <Phil C***@discussions.microsoft.com> wrote in message
news:1D95D27B-AD34-4666-9012-704E229BA585@microsoft.com...
> When I retrieve a 'float' type value in QA it sometimes displays as
> 2768.8400000000001 but when viewing the table in EM it shows as 2768.84,
> which I believe to be the correct value.
>
> What causes this?
>
>
> MSSQL 2000
> QA v 8.00.194

"Andrew J. Kelly" wrote:

> Float is an approximate value to begin with.  The gui that is displaying the
> value may round it off to suite it's needs and I believe that is what you
> are seeing.  If you want accuracy to that level use decimal and not float.
>
> --
> Andrew J. Kelly  SQL MVP
>
>
> "Phil Cook" <Phil C***@discussions.microsoft.com> wrote in message
> news:1D95D27B-AD34-4666-9012-704E229BA585@microsoft.com...
> > When I retrieve a 'float' type value in QA it sometimes displays as
> > 2768.8400000000001 but when viewing the table in EM it shows as 2768.84,
> > which I believe to be the correct value.
> >
> > What causes this?
> >
> >
> > MSSQL 2000
> > QA v 8.00.194
>
>
>

"Phil Cook" wrote:

> Thanks Andrew. I could understand if it was a rounding issue, but I'm pretty
> sure the value was only input with 2 dec places. How can the value have been
> changed?
>
>
>
> "Andrew J. Kelly" wrote:
>
> > Float is an approximate value to begin with.  The gui that is displaying the
> > value may round it off to suite it's needs and I believe that is what you
> > are seeing.  If you want accuracy to that level use decimal and not float.
> >
> > --
> > Andrew J. Kelly  SQL MVP
> >
> >
> > "Phil Cook" <Phil C***@discussions.microsoft.com> wrote in message
> > news:1D95D27B-AD34-4666-9012-704E229BA585@microsoft.com...
> > > When I retrieve a 'float' type value in QA it sometimes displays as
> > > 2768.8400000000001 but when viewing the table in EM it shows as 2768.84,
> > > which I believe to be the correct value.
> > >
> > > What causes this?
> > >
> > >
> > > MSSQL 2000
> > > QA v 8.00.194
> >
> >
> >

"Mike Epprecht (SQL MVP)" wrote:

> Hi
>
> float is an approximate data type. Not all values can be held correctly.
>
> In your case, 2768.84 gets stored as 2768.8400000000001 as it is not
> possible to represent 2768.8400000000000 in the float data type.
>
> As Andrew said, use Decimal data type if you don't want SQL Server to store
> a 'close enough' value.
>
> Regards
> Mike
>
> "Phil Cook" wrote:
>
> > Thanks Andrew. I could understand if it was a rounding issue, but I'm pretty
> > sure the value was only input with 2 dec places. How can the value have been
> > changed?
> >
> >
> >
> > "Andrew J. Kelly" wrote:
> >
> > > Float is an approximate value to begin with.  The gui that is displaying the
> > > value may round it off to suite it's needs and I believe that is what you
> > > are seeing.  If you want accuracy to that level use decimal and not float.
> > >
> > > --
> > > Andrew J. Kelly  SQL MVP
> > >
> > >
> > > "Phil Cook" <Phil C***@discussions.microsoft.com> wrote in message
> > > news:1D95D27B-AD34-4666-9012-704E229BA585@microsoft.com...
> > > > When I retrieve a 'float' type value in QA it sometimes displays as
> > > > 2768.8400000000001 but when viewing the table in EM it shows as 2768.84,
> > > > which I believe to be the correct value.
> > > >
> > > > What causes this?
> > > >
> > > >
> > > > MSSQL 2000
> > > > QA v 8.00.194
> > >
> > >
> > >