>>>>>Hi all,
>>>>>TIA
>>>>Perry,
>>>>Sounds like a SQL counting by grouping :select Field1, Field2, Field3, cnt(*) ;
>>>> from mytable ;
>>>> group by 1,2,3 ;
>>>> into array myarray && into cursor myCursor more parctical ?Cetin
>>>
>>>Don't you need to precede the group by with the order by? Otherwise the un-ordered results will be inaccurate.
>>Dan,
>>As long as "group by" and "order by" order sequences are same you don't need. "Group by" makes an implicit "order by". Including "order by" is needed if you want to override ordering sequence.select * from mytable ;
>> group by field1, field2, field3 ;
>> order by field1, field2, field3
>>* Is equivalant to
>>select * from mytable ;
>> group by field1, field2, field3
>>* In case override needed
>>select * from mytable ;
>> group by field1, field2, field3 ;
>> order by field3, field1, field2The real problem is to prevent "group by" doing implicit "order by" on sequential (mean each group is contiguous) but not sorted data.
>>Cetin
>
>In the following case as well?
>
>...
>order by fld1, fld2, fld3, fld4
>group by fld1, fld2, fld3
>...
Sure since they're different, it's type of override too.
Cetin