Re: Normalisation Problem

Hi
Yes it is normalized iff the key is made of the first two columns.

Don't denormalize too much you may end up with a big number of tables!

André

Vadim Tropashko wrote:

> In article <Ux7z6.2993$jz.252464_at_www.newsranger.com>, Aloha Kakuikanu says...
> >
> >We are working on Ariphmetics database for American Mathematical Society and
> >disagree upon the following problem in Schema Design. In my design I have a
> >table:
> >
> >Table ADD_SUBTRACT_FACTS
> >+----------+-----------+-----+------+
> >| OperandA | OperandB | Sum | Diff |
 +----------+-----------+------------+
> >| 1 | 1 | 2 | 0 |
> >| 5 | 3 | 8 | 2 |
> >| ... | ... | ... | ... |
> >+----------+-----------+-----+------+
> >
> >But other members of the team say that it's not normalized, and, therefore, have
> >to be splitted into two:
> >
>
> I would suggest to decompose Operands in your design into:
>
> OperandAinOnes
> OperandAinTens
> OperandAinHundreds
>
> so that your first column become:
>
> +----------------+----------------+--------------------+
> | OperandAinOnes | OperandAinTens | OperandAinHundreds |

 +----------------+----------------+--------------------+


> |       1        |       0        |          0         |


> |       5        |       0        |          0         |