· What are some good SAS programming practices for
processing very large data sets?
Answers were Sorted based on User's Feedback
Answer / sibin
Use
1.WHERE Condition
2.DROP KEEP in Set Statement
3.ARRAY
4.Mutually Exclusive Conditions like DO ,DO UNTIL,DO WHILE
Is This Answer Correct ? | 16 Yes | 1 No |
Answer / pricil kurian
1) Compressing dataset
2) Using where instead of if as subsetting conditions
3) Deleting the temporary datasets at the end of the program
4) If we use "if..else ..if conditions"..use most occuring
situations first
5)Use keep and drop appropriately to avoid the unneceesary
variable
6).Creattion of index variable
7)Use of array statement
8)Create one byte required numeric variable which we don't
use for calcualting process as charcater..Using this this
variable need 1 byte where numeric variable's minimum
length is 3.
9)Minimize the CPU time that SAS uses by executing the
minimum number of statements in the most efficient
order.
10)Use SQL for code simplification. A single SQL statement
is generally easier to code and understand than a
long series of SORT and DATA steps.
11)Use of data _null_; when we don't need to output into a
new dataset.
12)Use macros for redundant code.
13)When data sets are large and unsorted, the SQL inner
join may beter than SORT and MERGE.
Is This Answer Correct ? | 8 Yes | 0 No |
Answer / vicky
1.we can compress the dataset
2.if we wana do merge a bigdate set with another dataset
having few variables for instance a fact with dimension
having key variable and description column
in this case sorting n merging this bigdateset(fact) with
dimension apply formats.so dat u can avoid sort n merge.
Is This Answer Correct ? | 2 Yes | 1 No |
Answer / dvsr
1)compress the datasets
2)use DROP KEEP options while creating temp data sets
3)Index creation on the Frequently using variables
Is This Answer Correct ? | 0 Yes | 1 No |
what other SAS features do you use for error trapping and data validation?
Can you suggest us materials for sdtm mapping?
What is the good sas programming practices for processing large data sets?
what are the validation tools in sas?
Is the physical structure of the data set in the same orientation as the report? Do you need to reshape the data sets? What method should you use to reshape the data–DATA steps,PROC TRANSPOSE,output data set from a procedure?
how can u join the two tables without using proc sql Joins and nested queries ?
SAS System ?
HOW DO WE CREATE A SAS STORED PROCESS IN SAS EG ?
How can you limit the variables written to output dataset in data step?
How do you test for missing values?
what is fact table and factless table?
if a variable contain dates like "2015/01"---"2015/12" (yymm) ,How to add day to those dates,if them month is jan then 31 if the month is feb then 28 so on ...