Re: del dup recs

Posted in 1994

Paul Roberts — 27 May 1994 06:33:53 GMT — source: Informix-list mailing list archive (1991-1998)

In article <2s41io$idb@emory.mathcs.emory.edu> EAPC-AMO2@emh7.korea.army.mil writes: > >Problem: Duplicate data in (supposedly) unique columns. > >Question: How do I remove the records with duplicate ssn columns if my data >file has about 25,00 recs of which 8,000 ssn's are dups (according to a select >count(distinct ssn), but data in the other columns may be different? > >i.e. > +----------+---------+--------------+ > |name |ssn |address | > +----------+---------+--------------+ > |fred |121212121|who cares | > |fred Jr. |121212121|I sure don't | > |tony |131313131|don't know | > +----------+---------+--------------+ Lines: 6 Distribution: world Message-ID: <2s448h$s97@infmx.informix.com> References: <2s41io$idb@emory.mathcs.emory.edu> NNTP-Posting-Host: infmx.informix.com Well .. in this case do you want to delete BOTH fred AND fred Jr, or only one of them? If only one, do you any criteria to decide which one? Given that the data in the other columns is not duplicate, it is going to make a difference which one you choose. Paul