00 Replace Missing Values withing Medians from each sub-groups

Prev: proc power question for McNemar test
Next: proc power question for McNemar test

From: Sdoof on 19 Dec 2009 12:52

---

Dear SAS Listers,

I want to replace the missing values in my dataset with the
corresponding median across each group define by another class
variable: CountryCode
The dataset looks like this

CountryCode COMPANYAGE
UK 10
UK .
UK 12
FR 10
FR .
US 20
US .
US 10


I am using the following code but it is not doing what I want )i.e. it
replaces all missing values with the median of the last country (US in
the avove example)

proc univariate data=DB;
var
CompanyAge
;
class
CountryCode
;
output out=
decile_CompanyAge_
pctlpts=50
pctlpre=
CompanyAge_
;
run;

data DB;
set DB;
if(_n_ eq 1) then set decile_CompanyAge_;
by
CountryCode;
retain
CompanyAge_50;
;
run;

data DB;
set DB;
by CountryCode;
if missing(CompanyAge) THEN CompanyAge=CompanyAge_50;
run;

Thanks in advance :-)

From: data _null_; on 20 Dec 2009 09:03

---

On Dec 19, 11:52 am, Sdoof <olivier.vanpa...(a)gmail.com> wrote:
> Dear SAS Listers,
>
> I want to replace the missing values in my dataset with the
> corresponding median across each group define by another class
> variable: CountryCode
> The dataset looks like this
>
> CountryCode      COMPANYAGE
> UK                    10
> UK                      .
> UK                    12
> FR                    10
> FR                      .
> US                    20
> US                      .
> US                    10
>
> I am using the following code but it is not doing what I want )i.e. it
> replaces all missing values with the median of the last country (US in
> the avove example)
>
> proc univariate data=DB;
> var
> CompanyAge
> ;
> class
> CountryCode
> ;
> output out=
> decile_CompanyAge_
> pctlpts=50
> pctlpre=
> CompanyAge_
> ;
> run;
>
> data DB;
> set DB;
> if(_n_ eq 1) then set decile_CompanyAge_;
> by
> CountryCode;
> retain
> CompanyAge_50;
> ;
> run;
>
> data DB;
> set DB;
> by CountryCode;
>   if missing(CompanyAge) THEN CompanyAge=CompanyAge_50;
>   run;
>
> Thanks in advance :-)

Try this, it is much easier...

data have;
input CountryCode:$2. COMPANYAGE @@;
cards;
UK 10 UK . UK 12
FR 10 FR .
US 20 US . US 10
;;;;
run;
proc stdize reponly missing=median;
by notsorted CountryCode;
var Companyage;
run;
proc print;
run;

 | 
Pages: 1
Prev: proc power question for McNemar test
Next: proc power question for McNemar test