topic Re: number of observations read by SAS in SAS Procedures

number of observations read by SAS

marline — Sun, 02 Feb 2020 20:27:55 GMT

I inserted my data into the SAS, but it is not reading my total number of observations?

sorry, I am a beginner.

what could be the reason?

Re: number of observations read by SAS

Kurt_Bremser — Mon, 03 Feb 2020 01:41:15 GMT

What kind of data, and how did you "insert" it? Please post the log from the code you ran, using the {i} button.

Re: number of observations read by SAS

marline — Thu, 06 Feb 2020 10:31:13 GMT

data sickdays;
input days@@;
datalines;

8	10	5	0	6	9	5	15		
5	4	3	2	0	4	15
;
run
;
proc means n mean std stderr clm ; var days;
run;
proc ttest  h0=<9.4; var days ; run;

Re: number of observations read by SAS

ed_sas_member — Thu, 06 Feb 2020 10:35:10 GMT

Hi @marline

It seems that the values are separated by a tab.

You can specify it by adding an INFILE statement. Dlm = "09"x means that the delimiter is a tab (not as space or a comma for example).

data sickdays;
	infile datalines dlm="09"x;
	input days @@;
	datalines;
8	10	5	0	6	9	5	15		
5	4	3	2	0	4	15
;
run;

Re: number of observations read by SAS

marline — Thu, 06 Feb 2020 10:35:32 GMT

in the note section I get the following: 2 observations are read from your data.

Although all my friends get the result without any problem on their SAS 9.4, is the sas university edition needs a different coding especially in entering the data ?

because I faced the same problem with a set of data copied and pasted from the excel sheet, it was 30 observations and I got only 4 read ?

Re: number of observations read by SAS

ed_sas_member — Thu, 06 Feb 2020 10:40:21 GMT

Hi @marline

I think there is also an issue in the PROC TTEST. Indeed, you can't specify an operator other than equal in H0.

However, you can tell SAS that you conduct an unilateral test by using the SIDE option:

proc ttest data=sickdays h0=9.4 side=u;
	var days;
run;

SIDE = U : Specifies upper one-sided tests, in which the alternative hypothesis indicates a mean greater than the null value

SIDE = L : Specifies lower one-sided tests, in which the alternative hypothesis indicates a mean less than the null value

SIDE = 2 : Bilateral

Hope this helps.

Best,

Re: number of observations read by SAS

marline — Thu, 06 Feb 2020 10:45:04 GMT

thanks a lot, I appreciate it but why it read 17 observations while I have 15 ?