topic Re: idea about conditional statement in SAS Programming

idea about conditional statement

mona4u — Wed, 30 May 2018 15:39:33 GMT

Hi,

I have this dataset and this is the result I desire

Task_Name	Assigned_To	result	count
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	vpandya@clinical.com	Image Preview
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	vpandya@clinical.com	Image Preview
Photo Image Preview QC	tcollins@clinical.com	Photo Image Preview QC
Admin QC	araju@clinical.com	Admin QC
Image Preview	arodriguez@clinical.com	Image Preview
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	vpandya@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	araju@clinical.com	Admin QC
Image Preview	arodriguez@clinical.com	Image Preview
Admin QC	marnofsky@wcclinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	vpandya@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	vpandya@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	araju@clinical.com	Admin QC
Image Preview	arodriguez@clinical.com	Image Preview
Admin QC			1`
Admin QC	marnofsky@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
irRC RECIST Oncology	rbht@bidmc.harvard.edu
Post Pre-Derm QC	cwalsh@clinical.com	Post Pre-Derm QC
Post Review QC 1	araju@clinical.com	Post Review QC 1
Post Review QC 2	araju@clinical.com	Post Review QC 2
Pre-Dermatology	cwalsh@clinical.com	Pre-Dermatology
Radiology Review 1	csado@partners.org
Radiology Review 2	vassah@hotmail.com
irRC RECIST Oncology

and I want an idea to get the result I want

If the scan if scan(assigned_to,2,"@")="clinical.com" then result=assigned_to;

I want to create a variable that has the values that associated with the clinical.com email from the variable assigned_to

but also I want to create a variable called count witch count the blanks in the assigned_to variable if the value is in the result variable

Re: idea about conditional statement

mkeintz — Wed, 30 May 2018 15:16:35 GMT

Could you

Put the incoming data in the form of a sas data set, and
(really important) show what you want the resulting data set to look like.

Re: idea about conditional statement

ballardw — Wed, 30 May 2018 15:22:21 GMT

Suggestion would be to cut the example data by maybe half.

Then show the desired result as well, which should be easier with the reduced set.

Reason for showing the desired result is the phrase "has the values that associated with the clinical.com email" is not exactly clear.

Does this mean that you want to do something with the "task_name" variable? Get the user name from such as tcollins from the string tcollins@clinical.com or something else.

Since each record could only have 1 "blank" what you actually want to count appears to be something else, possibly a summary after this data set is created.

Re: idea about conditional statement

mona4u — Wed, 30 May 2018 15:24:48 GMT

I just edited my post.

this is a sample data has only one blank but my real data has more

Re: idea about conditional statement

Reeza — Wed, 30 May 2018 15:28:52 GMT

Why not use FIND/INDEX instead of SCAN, since you're searching for a specific string set.

Re: idea about conditional statement

mona4u — Wed, 30 May 2018 15:37:03 GMT

This point it doesn't really matter

Re: idea about conditional statement

Reeza — Wed, 30 May 2018 16:23:16 GMT

if find('@clincial.com', assigned_to, 'i')>0 then result=task_name;

Isn't that all you need?

Re: idea about conditional statement

mona4u — Wed, 30 May 2018 16:24:01 GMT

edited it

Re: idea about conditional statement

mona4u — Wed, 30 May 2018 16:55:34 GMT

I don't have a problem with this part
I have a problem with the variable count
I need to create count to count how many blanks are they if the values in the variable result without specifying the values of the variable result each time I run the program

Re: idea about conditional statement

Reeza — Wed, 30 May 2018 17:02:33 GMT

IF/THEN/DO.

Adding on to my previous answer, change it to conditional so that you can add the count. You only show one value of 1, so not sure why the rest are blank or if you want them incremented or something. If so, use RETAIN. If you cannot get it working at this point, post your code.

if <condition> then do;

<multiple sas statements>;

count=1;

end;

Please be much more clear in your initial question.

http://documentation.sas.com/?docsetId=lestmtsref&docsetTarget=n0el0y2a02ab1ln1pks3gbac1en3.htm&docsetVersion=9.4&locale=en

Re: idea about conditional statement

mona4u — Wed, 30 May 2018 17:11:18 GMT

I know if then do
the question is
if then always need to specify a value of the variable to subset
I want to specify the variable that has all those values but without mentioning the values

Re: idea about conditional statement

Reeza — Wed, 30 May 2018 18:06:50 GMT

@mona4u wrote:
I know if then do
the question is
if then always need to specify a value of the variable to subset
I want to specify the variable that has all those values but without mentioning the values

I have no idea what that means.

Re: idea about conditional statement

mona4u — Wed, 30 May 2018 18:11:10 GMT

I mean usually, we subset this way
if var=a and var1=" " then count
if var in ("a", "b") and var1=" " then count

but the code that I need is
if the value in the var then count without specifying the value of the variable

Re: idea about conditional statement

Reeza — Wed, 30 May 2018 18:12:35 GMT

Yeah, still not following. Show me 5 lines of data and exactly what you want as output - as two different data sets.

@mona4u wrote:
I mean usually, we subset this way
if var=a and var1=" " then count
if var in ("a", "b") and var1=" " then count

but the code that I need is
if the value in the var then count without specifying the value of the variable

Re: idea about conditional statement

mona4u — Wed, 30 May 2018 18:15:11 GMT

dataset

Task_Name	Assigned_To
Admin QC	tcollins@clinical.com
Image Preview	vpandya@clinical.com
Admin QC	tcollins@clinical.com
Image Preview	tcollins@clinical.com
Admin QC	tcollins@clinical.com
Image Preview	vpandya@clinical.com
Photo Image Preview QC	tcollins@clinical.com
Admin QC	araju@clinical.com
Image Preview	arodriguez@clinical.com
Admin QC	tcollins@clinical.com
Image Preview	tcollins@clinical.com
Admin QC	vpandya@clinical.com
Image Preview	tcollins@clinical.com
Admin QC	araju@clinical.com
Image Preview	arodriguez@clinical.com
Admin QC	marnofsky@wcclinical.com
Image Preview	tcollins@clinical.com
Admin QC	vpandya@clinical.com
Image Preview	tcollins@clinical.com
Admin QC	tcollins@clinical.com
Image Preview	tcollins@clinical.com
Admin QC	vpandya@clinical.com
Image Preview	tcollins@clinical.com
Admin QC	araju@clinical.com
Image Preview	arodriguez@clinical.com
Admin QC
Admin QC	marnofsky@clinical.com
Image Preview	tcollins@clinical.com
irRC RECIST Oncology	rbht@bidmc.harvard.edu
Post Pre-Derm QC	cwalsh@clinical.com
Post Review QC 1	araju@clinical.com
Post Review QC 2	araju@clinical.com
Pre-Dermatology	cwalsh@clinical.com
Radiology Review 1	csado@partners.org
Radiology Review 2	vassah@hotmail.com
irRC RECIST Oncology

result I desire

Task_Name	Assigned_To	result	count
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	vpandya@clinical.com	Image Preview
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	vpandya@clinical.com	Image Preview
Photo Image Preview QC	tcollins@clinical.com	Photo Image Preview QC
Admin QC	araju@clinical.com	Admin QC
Image Preview	arodriguez@clinical.com	Image Preview
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	vpandya@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	araju@clinical.com	Admin QC
Image Preview	arodriguez@clinical.com	Image Preview
Admin QC	marnofsky@wcclinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	vpandya@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	tcollins@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	vpandya@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
Admin QC	araju@clinical.com	Admin QC
Image Preview	arodriguez@clinical.com	Image Preview
Admin QC			1`
Admin QC	marnofsky@clinical.com	Admin QC
Image Preview	tcollins@clinical.com	Image Preview
irRC RECIST Oncology	rbht@bidmc.harvard.edu
Post Pre-Derm QC	cwalsh@clinical.com	Post Pre-Derm QC
Post Review QC 1	araju@clinical.com	Post Review QC 1
Post Review QC 2	araju@clinical.com	Post Review QC 2
Pre-Dermatology	cwalsh@clinical.com	Pre-Dermatology
Radiology Review 1	csado@partners.org
Radiology Review 2	vassah@hotmail.com
irRC RECIST Oncology

Re: idea about conditional statement

Reeza — Wed, 30 May 2018 20:13:59 GMT

Why isn't the last record included in the count?

Re: idea about conditional statement

mona4u — Wed, 30 May 2018 20:23:01 GMT

I only count the blanks in assigned_to
if the value of task was in the result
"irRC RECIST Oncology" is not one of the variable result's values

Re: idea about conditional statement

Reeza — Wed, 30 May 2018 20:54:58 GMT

if the value of task was in the result

How is this identified? Do you have a list somewhere that identifies what's valid and what's not?

Re: idea about conditional statement

mona4u — Thu, 31 May 2018 15:28:27 GMT

I don't have a list
the idea is I should include whatever is task_name if the email has "clinical.com"
I want the program to be able to create a list automatically and then
check the blanks assigned_to if anything of the list is there

Re: idea about conditional statement

Reeza — Thu, 31 May 2018 15:35:25 GMT

Create your list then first:

proc sql noprint;
select distinct quote(task_name) into :task_list separated by ", "
from have
where email like '%@clinical.com';
quit;
%put &task_list;

If the macro variable is too long you'll need a different approach.