Question

This seems like it should be easier than I’m finding it. I have a table that contains both first and last names (specified by a type ID) and a frequency of how common the name is. For example:

NameType | Name      | FrequencyPercent
1          John        3.267
1          Thomas      1.987
1          Wilson      0.945
2          Smith       4.528
2          Wilson      2.221
2          Thomas      0.437

I want to be able to query out a list of First Names (NameType = 1) that will only include the name if it is more likely to be a First Name than a Last Name (based on the FrequencyPercentage). With this dataset, my FirstNames query would include John and Thomas, and my LastNames query would return Smith and Wilson.

Hopefully I explained that well.

Thanks ahead for a little help,

Russell Schutte

Answer 1

If I understand correctly you’re looking for first names when the frequency is higher than the frequency as same name as last name

This works for first names. You just need to reverse it for last names

CREATE Table YourTable
(
NameType int,
name varchar(20),
FrequencyPercent decimal(12,4)
)

INSERT INTO  YourTable
VALUES (1 ,'John', 3.267),
(1 , 'Thomas',      1.987),
(1 , 'Wilson',      0.945),
(2 , 'Smith',       4.528),
(2 ,  'Wilson',      2.221),
(2 ,   'Thomas',      0.437)

SELECT firstNames.name
FROM
      YourTable firstNames 
LEFT JOIN YourTable  lastNames 
 ON firstnames.Name = lastNames.Name
    AND lastNames.NameType  =2
     and firstnames.FrequencyPercent < lastNames.FrequencyPercent
WHERE firstNames.NameType  =1
      AND
      lastNames.name is null

results in

name
--------------------
John
Thomas
(2 row(s) affected)