topic Re: How do I add known constants as a column to a dataset in New SAS User

How do I add known constants as a column to a dataset

heyyou1 — Tue, 16 Nov 2021 20:29:09 GMT

I have a data set that I want to add known constants as a new column to the dataset. The constants are not likely to change.

data money;
	infile datalines delimiter=",";
	input name $ return invested;
	datalines;
Joe,10,100
Bob,7,50
Mary,80,1000
;

this is the existing dataset, and I now want to add the column `height`. (I know this is silly but its for the question). My current attempt is this:

data returns;
    set returns;
        heights(m) = 1.8, 2.1, 1.6;
run;

The end result I want is something like this.

/* name     | return | invested | height(m)*/
/* ____________________________________________ */
/* Joe      |  10   | 100       |  1.8 */
/* Bob      |   7   | 50        | 2.1 */
/* Mary     | 80    | 50        | 1.6 */

how could I do something like that?

Re: How do I add known constants as a column to a dataset

Reeza — Tue, 16 Nov 2021 20:35:22 GMT

You would rarely ever process data like that manually. Can you explain your actual situation? Would the data come from a data set, sql, JSON?

Re: How do I add known constants as a column to a dataset

heyyou1 — Tue, 16 Nov 2021 20:40:05 GMT

The dataset is coming from a data set about the surrounding population and their health status. The constant data is the total population estimate by race. The total population estimates by race can be found on google and are used for calculating percentages within that data set.

Re: How do I add known constants as a column to a dataset

Reeza — Tue, 16 Nov 2021 20:55:37 GMT

Personally, this is what I'd do:

data heights;
	infile datalines delimiter=",";
	input name $ height;
	datalines;
Joe,1.8
Bob,2.1
Mary,1.6
;

Then merge it by name, not a fan of implicit merges based on row positions.

proc sort data=money; by name;
proc sort data=heights; by name;

data combined;
merge money heights;
by name;
run;

@heyyou1 wrote:
The dataset is coming from a data set about the surrounding population and their health status. The constant data is the total population estimate by race. The total population estimates by race can be found on google and are used for calculating percentages within that data set.

Re: How do I add known constants as a column to a dataset

Tom — Tue, 16 Nov 2021 21:08:22 GMT

@heyyou1 wrote:
The dataset is coming from a data set about the surrounding population and their health status. The constant data is the total population estimate by race. The total population estimates by race can be found on google and are used for calculating percentages within that data set.

Let's create an example that more closely matches this description than the code in your original question.

So you have some data that includes RACE as one of the variables.

data have;
  input race :$40. var1 var2;
cards;
....

And you have your population data by race.

data population;
  input race :$40. total;
cards;
....

You then merge it with your dataset that has the population by RACE.

Now you have the population number on every observation of your original data (at least the ones that have race categories that match).

data want;
  merge have population;
  by race;
  calculation1 = var1/total;
run;