I'm attempting to use PROC SQL to sum population counts (i.e., whole numbers) for various age groups and counties:
PROC SQL;
CREATE TABLE WORK.MALE_POP_SQL00 AS
SELECT countyname AS CountyName, age_grp AS AgeGroup, SUM(pop00) AS Pop_00
FROM WORK.INTERCENSAL_M
GROUP BY countyname, age_grp
ORDER BY countyname, age_grp;
QUIT;
The issue I'm having is that the results given aren't whole numbers:
Results
Worse than that is that rounding often produces counts that don't match their original raw values. For example the last entry - the raw population value for that age group / county is actually 4, but after rounding the value produced by the PROC SQL it's 3.
Thanks for your time.
You most likely don't actually have whole numbers in your data, despite thinking you do. Formats can hide decimals, but they don't change the actual value, and SQL will not preserve the format except in direct select statements with no grouping/etc.
See for example:
data my_data;
input age_grp pop_count;
format pop_count 4.0;
datalines;
1 1234.54
2 1624.43
1 14.35
2 1234.11
1 888.88
2 768.48
;;;;
run;
proc sql;
select age_grp, sum(pop_count) as pop_sum
from my_data
group by age_grp;
quit;
And age_grp 2 does not add up rounded to the same thing as it adds to unrounded (the rounded values appear to add up to 3626).
Related
I have a dataset with 20 columns all starting with the name morb_, which are all 1 or 2, coded as No and Yes. There is an additional column called Pat_TNO which is the patient reference number. Patients have more than one row.
I wish to create a new dataset which summarises whether each patient has had at least one of each type of event. So far the code I have written works perfectly, but is there a way to simplify it using an array?
proc sql;
select
Pat_TNO,
max(morb_1) as morb_1 format yn.,
max(morb_2) as morb_2 format yn. /* etc etc */
from morbidity
group by Pat_TNO;
quit;
COumn names aren't morb_1 and morb_2, rather morb_amputation, morb_mi, morb_tia, etc.
proc summary data=morbidity nway missing;
class pat_tno;
output out=max max(morb_:) = ;
run;
Dataset HAVE includes id values and a character variable of names. Values in names are usually missing. If names is missing for all values of an id EXCEPT one, the obs for IDs with missing values in names can be deleted. If names is completely missing for all id of a certain value (like id = 2 or 5 below), one record for this id value must be preserved.
In other words, I need to turn HAVE:
id names
1
1
1 Matt, Lisa, Dan
1
2
2
2
3
3
3 Emily, Nate
3
4
4
4 Bob
5
into WANT:
id names
1 Matt, Lisa, Dan
2
3 Emily, Nate
4 Bob
5
I currently do this by deleting all records where names is missing, then merging the results onto a new dataset KEY with one variable id that contains all original values (1, 2, 3, 4, and 5):
data WANT_pre;
set HAVE;
if names = " " then delete;
run;
data WANT;
merge KEY
WANT_pre;
by id;
run;
This is ideal for HAVE because I know that id is a set of numeric values ranging from 1 to 5. But I am less sure how I could do this efficiently (A) on a much larger file, and (B) if if I couldn't simply create an id KEY dataset by counting from 1 to n. If your HAVE had a few million observations and your id values were more complex (e.g., hexadecimal values like XR4GN), how would you produce WANT?
You can use SQL here easily, MAX() applies to character variables within SQL.
proc sql;
create table want as
select id, max(names) as names
from have
group by ID;
quit;
Another option is to use an UPDATE statement instead.
data want;
update have (obs=0) have;
by ID;
run;
This seems like a good candidate for a DOW-loop, assuming that your dataset is sorted by id:
data want;
do until(last.id);
set have;
by id;
length t_names $50; /*Set this to at least the same length as names unless you want the default length of 200 from coalescec*/
t_names = coalescec(t_names,names);
end;
names = t_names;
drop t_names;
run;
proc summary data=have nway missing;
class id;
output out=want(drop=_:) idgroup(max(names) out(names)=);
run;
Use the UPDATE statement. That will ignore the missing values and keep the last non-missing value. It normally requires a master and transaction dataset, but you can use your single dataset for both.
data want;
update have(obs=0) have ;
by id;
run;
How can I convert my SAS data set, into a data set that I can easily paste into the forum or hand over to someone to replicate my data. Ideally, I'd also like to be able to control the amount of records that are included.
Ie I have sashelp.class in the SASHELP library, but I want to provide it here so others can use it as the starting point for my question.
To do this, you can use a macro written by Mark Jordan at SAS, the code is stored in GitHub as well.
You need to provide the data set name, including library and the number of observations you want to output. It takes them in order. The code will then appear in your SAS log.
*data set you want to create demo data for;
%let dataSetName = sashelp.Class;
*number of observations you want to keep;
%let obsKeep = 5;
******************************************************
DO NOT CHANGE ANYTHING BELOW THIS LINE
******************************************************;
%let source_path = https://gist.githubusercontent.com/statgeek/bcc55940dd825a13b9c8ca40a904cba9/raw/865d2cf18f5150b8e887218dde0fc3951d0ff15b/data2datastep.sas;
filename reprex url "&source_path";
%include reprex;
filename reprex;
option linesize=max;
%data2datastep(dsn=&dataSetName, obs=&obsKeep);
This may not work if you do not have access to the github page, in that case, you can manually navigate to the page (same link) and copy/paste it into SAS. Then run the program and run only the last step, the %data2datastep(dsn=, obs=);
This topic came up recently on SAS Communities and I created a little more robust macro than the one Reeza linked. You can see it in Github: ds2post.sas
* Pull macro definition from GITHUB ;
filename ds2post url
'https://raw.githubusercontent.com/sasutils/macros/master/ds2post.sas'
;
%include ds2post ;
For example if you wanted to share the first 5 observations of SASHELP.CARS you would run this macro call:
%ds2post(sashelp.cars,obs=5)
Which would generate this code to the SAS log:
data work.cars (label='2004 Car Data');
infile datalines dsd dlm='|' truncover;
input Make :$13. Model :$40. Type :$8. Origin :$6. DriveTrain :$5.
MSRP Invoice EngineSize Cylinders Horsepower MPG_City MPG_Highway
Weight Wheelbase Length
;
format MSRP dollar8. Invoice dollar8. ;
label EngineSize='Engine Size (L)' MPG_City='MPG (City)'
MPG_Highway='MPG (Highway)' Weight='Weight (LBS)'
Wheelbase='Wheelbase (IN)' Length='Length (IN)'
;
datalines4;
Acura|MDX|SUV|Asia|All|36945|33337|3.5|6|265|17|23|4451|106|189
Acura|RSX Type S 2dr|Sedan|Asia|Front|23820|21761|2|4|200|24|31|2778|101|172
Acura|TSX 4dr|Sedan|Asia|Front|26990|24647|2.4|4|200|22|29|3230|105|183
Acura|TL 4dr|Sedan|Asia|Front|33195|30299|3.2|6|270|20|28|3575|108|186
Acura|3.5 RL 4dr|Sedan|Asia|Front|43755|39014|3.5|6|225|18|24|3880|115|197
;;;;
Try this little test to compare the two macros.
First make a sample dataset with a couple of issues.
data testit;
set sashelp.class (obs=5);
if _n_=1 then name='Le Bron';
if _n_=2 then age=.;
if _n_=3 then wt=.;
if _n_=4 then name='12;34';
run;
Then run both macros to dump code to the SAS log.
%ds2post(testit);
%data2datastep(dsn=testit,obs=20);
Copy the code from the log. Changing the name in the DATA statements to not overwrite the original dataset or each other. Run them and compare the result to the original.
proc compare data=testit compare=testit1; run;
proc compare data=testit compare=testit2; run;
Result using %DS2POST:
The COMPARE Procedure
Comparison of WORK.TESTIT with WORK.TESTIT1
(Method=EXACT)
Data Set Summary
Dataset Created Modified NVar NObs
WORK.TESTIT 02NOV18:17:09:40 02NOV18:17:09:40 6 5
WORK.TESTIT1 02NOV18:17:10:29 02NOV18:17:10:29 6 5
Variables Summary
Number of Variables in Common: 6.
Observation Summary
Observation Base Compare
First Obs 1 1
Last Obs 5 5
Number of Observations in Common: 5.
Total Number of Observations Read from WORK.TESTIT: 5.
Total Number of Observations Read from WORK.TESTIT1: 5.
Number of Observations with Some Compared Variables Unequal: 0.
Number of Observations with All Compared Variables Equal: 5.
Summary of results using %Data2DataStep:
Comparison of WORK.TESTIT with WORK.TESTIT2
(Method=EXACT)
Data Set Summary
Dataset Created Modified NVar NObs
WORK.TESTIT 02NOV18:17:09:40 02NOV18:17:09:40 6 5
WORK.TESTIT2 02NOV18:17:10:29 02NOV18:17:10:29 6 3
Variables Summary
Number of Variables in Common: 6.
Observation Summary
Observation Base Compare
First Obs 1 1
First Unequal 1 1
Last Unequal 3 3
Last Match 3 3
Last Obs 5 .
Number of Observations in Common: 3.
Number of Observations in WORK.TESTIT but not in WORK.TESTIT2: 2.
Total Number of Observations Read from WORK.TESTIT: 5.
Total Number of Observations Read from WORK.TESTIT2: 3.
Number of Observations with Some Compared Variables Unequal: 3.
Number of Observations with All Compared Variables Equal: 0.
Variable Values Summary
Values Comparison Summary
Number of Variables Compared with All Observations Equal: 1.
Number of Variables Compared with Some Observations Unequal: 5.
Number of Variables with Missing Value Differences: 4.
Total Number of Values which Compare Unequal: 12.
Maximum Difference: 0.
Variables with Unequal Values
Variable Type Len Ndif MaxDif MissDif
Name CHAR 8 1 0
Sex CHAR 1 3 3
Age NUM 8 2 0 2
Height NUM 8 3 0 3
Weight NUM 8 3 0 3
Note that I am sure there are values that will cause trouble for my macro also. But hopefully they are caused by data that is less likely to occur than spaces or semi-colons.
I have the following data:
data df;
input id $ d1 d2 d3;
datalines;
a . 2 3
b . . .
c 1 . 3
d . . .
;
run;
I want to apply some transformation/operation across a subset of columns. In this case, that means dropping all rows where columns prefixed with d are all missing/null.
Here's one way I accomplished this, taking heavy influence from this SO post.
First, sum all numeric columns, row-wise.
data df_total;
set df;
total = sum(of _numeric_);
run;
Next, drop all rows where total is missing/null.
data df_final;
set df_total;
where total is not missing;
run;
Which gives me the output I wanted:
a . 2 3
c 1 . 3
My issue, however, is that this approach assumes that there's only one "primary-key" column (id, in this case) and everything else is numeric and should be considered as a part of this sum(of _numeric_) is not missing logic.
In reality, I have a diverse array of other columns in the original dataset, df, and it's not feasible to simply drop all of them, writing all of that out. I know the columns for which I want to run this "test" all are prefixed with d (and more specifically, match the pattern d<mm><dd>).
How can I extend this approach to a particular subset of columns?
Use a different short cut reference, since you know it all starts with D,
total = sum( of D:);
if n(of D:) = 0 then delete;
Which will add variables that are numeric and start with D. If you have variables you want to exclude that start with D, that's problematic.
Since it's numeric, you can also use the N() function instead, which counts the non missing values in the row. In general though, SAS will do this automatically for most PROCS such as REG/GLM(not in a data step obviously).
If that doesn't work for some reason you can query the list of variables from the sashelp table.
proc sql noprint;
select name into :var_list separated by ", " from sashelp.vcolumn
where libname='WORK' and memname='DF' and name like 'D%';
quit;
data df;
set have;
if n(&var_list.)=0 then delete;
run;
I'm very new to SAS and i'm trying to figure out my way around using it. I'm trying to figure out how to use the Compare procedure. Basically what I want to do is to see if the values in one column match the values in another column multiplied by 2 and count the number of mistakes. So if I have this data set:
a b
2 4
1 2
3 5
It should check whether b = 2 * a and tell me how many errors they are. I've been reading through the documentation for the compare procedure but like i said i'm very new and i can't seem to figure out how to check for this.
You could do if with PROC COMPARE but you still need to compute 2*a and you can't do that with PROC COMPARE. I would create a FLAG and summarize the FLAG. IFN function returns 1 for values that are NOT equal. PROC MEANS counts the 1's where mean is percent and sum is count of non-matching.
data comp;
input a b;
flag = ifn(b NE 2*a,1,0);
cards;
2 4
1 2
3 5
;;;;
run;
proc means n mean sum;
var flag;
run;
Proc compare compares values in two different datasets, whereas your variables are both in one dataset. The following may be simplest:
data matches errors;
set temp;
if b = 2 * a then output matches;
else output errors;
run;