SAS procedures those I have used

Proc Print Data=Example_Data;Run;

This procedure is very helpful for me when I want to glance the data to get a sense how do it look and to validate output of any code. This can be used with options like obs=N to print only top N observations. When firstobs=M option is used with   obs=N  (where N>M) along with the print procedure it prints observations from M to N. So you can view observations irrespective of   its order in the dataset. 

Proc expand:

 A year back I had used a procedure called Proc Expand to impute missing values in time series data and also to convert data from high level to lower level such as from week level to day level or low level to high level such as week to month. This procedure is very handy when we want to impute missing values in time series data of thousand of the line item.  In other words if we have panel data then we can use it.  Its syntax is very simple. It uses extrapolation and interpolation techniques to    convert the data from high to low level. But it when using this procedure you should be careful as it also results in negative values.   The syntax is written below

PROC EXPAND DATA=  options ;
BY variables ;
CONVERT variables / options ;
ID variable ;
Run;
By variables : It list the variables which indicates levels of a cross-sectional variable;
ID variables : indicates SAS time variable;
Convert Variable indicates the variable which has to be aggregated or interpolated. Using the option we can define the method to aggregate/interpolate
In the data option we can define the output data set name and frequency of aggregation/interpolation in the time dimension. Such year to quarter etc;


'

Comments

Popular posts from this blog

How to check whether a SAS dataset exist or not and throw an error in the log ?

Solution for ERROR: Some character data was lost during transcoding in the dataset

2018 plan for getting expertise in Machine Learning and Deep Learning