BookmarkSubscribeRSS Feed
Dont
Calcite | Level 5

 

Hello,

 

I have some data from groups of people spending money at a number of stores in a mall. I'm trying to calculate the percentage of people who spent at least 25 or more and at least 30 or more using proc report. I've attached the sas file and some of my code so far, but it is pretty basic. I assume I need to use sum and pctsum, but I don't how to incorporate them into the code correctly.

 

Any help is much appreciated

 

proc report data=mydata.house;
column id spent;
define id / display;
define spent / display;


title1 'Expenditure per group';

run;

 

5 REPLIES 5
ballardw
Super User

It is better to post data in the form of datastep code such as generated with this:  https://communities.sas.com/t5/SAS-Communities-Library/How-to-create-a-data-step-version-of-your-dat... will show how to turn an existing SAS data set into data step code that can be pasted into a forum code box using the {i} icon or attached as text to show exactly what you have and that we can test code against.

 

SAS data sets from different operating systems may not work, depending on the options in effect a 9.4 data set may not run with 9.2 or 9.3, 32-bit vs 64-bit, lanuage encoding may cause issues. With a data step everyone can run the code on their system to recreate a data set.

 

When you say:

" I'm trying to calculate the percentage of people who spent at least 25 or more and at least 30 or more using proc report."

is the 25 and 30 you reference a number of purchases (n) a total currency amount or some percent used as a group? When you mention percent you should clearly indicate what a numerator and denominator may be.

If you provide a small data set that you can calculate the result by hand then it is a very good idea to show what you want the output to look like. Otherwise we have to make assumptions and then you get to say "what I really wanted was xxxxx" so start with the XXXX spelled out.

HB
Barite | Level 11 HB
Barite | Level 11

You'll need to provide more context and perhaps some data.

 

"people who spent at least 25 or more"

 

I don't know what that means.  People who had spending at 25 stores?  People who spent $25? People who exhausted themselves (became spent) 25 times?

 

data spending;
   input spending;
datalines;
25
20
15
30
35
40
35
60
;
run;


PROC FORMAT; 
VALUE spending_fmt
	0-25  = "Spent $25 or less"
	26 - HIGH  = "Spent more than $25";
RUN; 

proc freq data=spending;
tables spending / missing;
FORMAT spending spending_fmt.;
run;
                                       The FREQ Procedure

                                                            Cumulative    Cumulative
                       spending    Frequency     Percent     Frequency      Percent
            
            Spent $25 or less             3       37.50             3        37.50
            Spent more than $25           5       62.50             8       100.00

 

Dont
Calcite | Level 5

Thank you very much for your input so far. I'm still working through how to use Mark Jordan's program so I've just posted an example of my code.

 

To clarify re: percentage, each ID represents a person. The 5 1s represent separate purchases that 1 person made.

 

i.e. person 1 bought 5 items of varying cost. Person 3 bought 4 items of varying cost.

 

I would like to calculate the percentage of people who spent more than 25 and more than 30  using proc report.

 

15.481
10.911
11.471
13.321
14.011
23.083
10.383
14.013
17.983
527.184
625.64
642.274
22.35
48.435
49.485
3.796
2.636
3.466
6.576
16.416

 

Thanks again

HB
Barite | Level 11 HB
Barite | Level 11

More than $25 in total or had at least one purchase exceeding $25? 100 percent of those ID's spent more than $25 in total. 20 percent of those ID's had at least one expenditure exceeding $25 dollars.

Which is the correct answer?

 

Edit:

 

Just for amusement.

 

data expenditures;
   input spending 8.2 id;
datalines;
15.48	1
10.91	1
11.47	1
25.32	1
14.01	1
23.08	3
10.38	3
14.01	3
17.98	3
527.18	4
625.6	4
642.27	4
22.3	5
48.43	5
49.48	5
3.79	6
2.63	6
3.46	6
6.57	6
16.41	6
;
run;

data flagged_expenditures;
	set expenditures;
	big_spender = 0;
	if spending >= 25 then big_spender = 1;
run;

proc sort data = flagged_expenditures;
	by id descending big_spender;
run;

data big_spenders;
	set flagged_expenditures;
	by id;
	if first.id;
run;

PROC FORMAT; 
VALUE big_spender_fmt
	0 = "Did not spend more than $25 on one purchase"
	1 = "Spent more than $25 on one purchase";
RUN; 

proc freq;
	tables big_spender;
	format big_spender big_spender_fmt.;
run;

yields

 

                                       The FREQ Procedure

                                                                        Cumulative    Cumulative
                                big_spender    Frequency     Percent     Frequency      Percent

Did not spend more than $25 on one purchase           2       40.00             2        40.00
Spent more than $25 on one purchase                   3       60.00             5       100.00

 

 

2nd edit:

 

And just becasue we can and it amuses me:

 

proc sql;
	create table big_spenders2 as
	select 'Big Spender' as status, count(a.id) as mycount
	from (
	select id, sum(case when spending > 25 then 1 else 0 end) as big_spender
	from expenditures
	group by id) as a
	where a.big_spender >0
	union
	select 'Little Spender' as status, count(a.id) as mycount
	from (
	select id, sum(case when spending > 25 then 1 else 0 end) as big_spender
	from expenditures
	group by id) as a
	where a.big_spender = 0;
quit;

proc sql;
	create table big_spender_percentage as 
	select status, mycount / (select sum(mycount) from big_spenders2) as mypercent
	from big_spenders2;
quit;

proc print;
run;

yeilds

                               Obs        status        mypercent

                                1     Big Spender          0.6
                                2     Little Spender       0.4

Dont
Calcite | Level 5

I'm interested in at least one purchase >25, who made 2 purchases>25 and then who made 3 purchases >25. After that I was planning on calculating it for different amounts (i.e. 30, 50 etc), but I was hoping I'd be able to do that myself.

 

Thank you very much for your help.

sas-innovate-2024.png

Join us for SAS Innovate April 16-19 at the Aria in Las Vegas. Bring the team and save big with our group pricing for a limited time only.

Pre-conference courses and tutorials are filling up fast and are always a sellout. Register today to reserve your seat.

 

Register now!

How to Concatenate Values

Learn how use the CAT functions in SAS to join values from multiple variables into a single value.

Find more tutorials on the SAS Users YouTube channel.

Click image to register for webinarClick image to register for webinar

Classroom Training Available!

Select SAS Training centers are offering in-person courses. View upcoming courses for:

View all other training opportunities.

Discussion stats
  • 5 replies
  • 976 views
  • 0 likes
  • 3 in conversation