topic Re: need help to extract middle name from heterogeneous variable with different words coun in SAS Programming

need help to extract middle name from heterogeneous variable with different words coun

mrahouma — Thu, 26 Sep 2019 20:37:24 GMT

I have the following dataset from which I need help to extract middle name from heterogeneous variable with different words count.

I saw prior link but it does not work for me.

my data has only 1 variable

Data statisticians;
infile datalines ;
Input name $30. ;
Datalines;
Ronaldo Al Fisher
H. O. Meir
Lee Sara Kim Ivan
Marco Sina
;
run;

Names with first and last names only should be blank in the new column called middle.

I tried the following code but it does not work:

data statisticians;  length middle $10; set statisticians; if
 count = 2 then middle=.; if count = 3 then middle= scan(name,2);   
if count = 4 then middle=scan(name,2);  run;

Any help will be greatly appreciated.

Re: need help to extract middle name from heterogeneous variable with different words coun

Kurt_Bremser — Thu, 26 Sep 2019 20:51:38 GMT

data statisticians;
infile datalines;
input name $30.;
length middle $10;
if countw(name) > 2 then middle = scan(name,2);
datalines;
Ronaldo Al Fisher
H. O. Meir
Lee Sara Kim Ivan
Marco Sina
;

Re: need help to extract middle name from heterogeneous variable with different words coun

Shmuel — Thu, 26 Sep 2019 20:55:49 GMT

I don't think there is a rule to define which string in a four string name is the real middle name.

If you can suggest such a rule then you can adapt the code you got.

Re: need help to extract middle name from heterogeneous variable with different words coun

mrahouma — Thu, 26 Sep 2019 21:01:34 GMT

Thanks a lot. It works

Re: need help to extract middle name from heterogeneous variable with different words coun

ballardw — Thu, 26 Sep 2019 21:04:44 GMT

@mrahouma wrote:

I have the following dataset from which I need help to extract middle name from heterogeneous variable with different words count.

I saw prior link but it does not work for me.

my data has only 1 variable

Data statisticians;
infile datalines ;
Input name $30. ;
Datalines;
Ronaldo Al Fisher
H. O. Meir
Lee Sara Kim Ivan
Marco Sina
;
run;

Names with first and last names only should be blank in the new column called middle.

I tried the following code but it does not work:
data statisticians;  length middle $10; set statisticians; if
 count = 2 then middle=.; if count = 3 then middle= scan(name,2);   
if count = 4 then middle=scan(name,2);  run;
Any help will be greatly appreciated.

Are you 100 percent sure that your data does not have an last names that are two or more words such as "Van Dyke" "De La Cruz" without middle names present?

Or possibly have only a last name (or first)?

Or have added bits like titles (Dr John Doe, Mrs Jane Roe) or indications like "John Smith II" or "John Smith Junior" or "John Smith the Third"?

You may want to bring anything with a count > 5 to personal attention.