Page 1 of 1
MatchFrequency file Vs. the weightage
Posted: Fri Mar 20, 2009 3:26 am
by vimalvik
How does the source volume affects the frequency file generation process and in turn how it affects the matching process?
Match frequency file is generated for 100m records and the source data is having some 200M.
Posted: Fri Mar 20, 2009 3:35 pm
by ray.wurlod
How are you generating the match frequencies? Are you perhaps on a two node environment and only looking at figures from one node?
Re: MatchFrequency file Vs. the weightage
Posted: Tue Mar 24, 2009 2:10 am
by vairus
Hi,
MatchFrequency file describe how often a value appears in source column.
If a name appears 100 times in a column. MatchFrequency file output for that name will be a single row with number of occurance and statistical weight.
So 200M records can generate lesser output records.
In Matching process , less weight will be given to the value which occured many times and more weight will be given to the value which occured few times.
Regards
vimalvik wrote:How does the source volume affects the frequency file generation process and in turn how it affects the matching process?
Match frequency file is generated for 100m records and the source data is having some 200M.