Arithmetic operations in sas

[#1]

Hi everyone.

I have a data set and I attached it to my message.  You can see it below.

I want to calculate bolt_volume. It looks simple but I couldn't get correct results. The formula has already written in excel sheet.

bolt_volume[row2] = ( bolt_area[row1] + bolt_area[row2] ) / 2 * bolt_len[row2]

bolt_volume[row3] = ( bolt_area[row2] + bolt_area[row3] ) / 2 * bolt_len[row3]

...........

tree_no	bolt_no	bolt_len	bolt_area	bolt_volume
1			1.235344635
1	1	2	0.89358336	2.128927995
1	2	2	0.821784859	1.715368219
2			0.346877809
2	1	2	0.276875719	0.623753528
2	2	2	0.226900035	0.503775754
3			0.294638715
3	1	2	0.193085235	0.48772395
3	2	2	0.185057629	0.378142864

Thank you for your help.

[#2]

Hi,

This should work for the desired output.

Thanks,

Naeem

data want;

set have;

bolt_volume=(bolt_area+lag(bolt_area))/2*bolt_len;

run;

[#3 · Accepted solution]

Here is one way:

data want;

  set have;

  by tree_no;

  bolt_volume=ifn(first.tree_no,.,lag(bolt_area)+bolt_area) / 2 * bolt_len;

run;

[#4]

Thanks stat@sas  and  Arthur Tabachneck ,

Both of your codes worked perfectly.

[#5]

Just be careful with Naeem's code as it will give you non-missing, and wrong values, for the 1st record of the 2nd thru Nth by groups.

[#6]

@Arthur Tabachneck,

My original data set has more than 2000 observations. After I saw your message, I checked your codes and Nadeem's codes for my data set. But I couldn't see wrong values 1st record of the 2nd through Nth by groups.

[#7]

I stand corrected!  I hadn't noticed that bolt_len was always missing for your first record within an id.  If that is the case in all situations, then Naeem's code will provide the correct result.

[#8]

Thanks Arthur - Yes, if first id is not missing then your suggested solution is a definite choice.

Regards,

Naeem

[#9]

Ok. I see now. That's tender spot.

Thanks Arthur.

[#10]

An operator in SAS is a symbol which is used in a mathematical, logical or comparison expression.

Example Arithmetic Operators.

DATA MYDATA1;
input @1 COL1 4.2 @7 COL2 3.1;
Add_result = COL1+COL2;
Sub_result = COL1-COL2;
Mult_result = COL1*COL2;
Div_result = COL1/COL2;
Expo_result = COL1**COL2;
datalines;
11.21 5.3
3.11 11
;
PROC PRINT DATA=MYDATA1;
RUN;

On running the above code, we get the following output.