PROC APPEND - appending data vertically

PROC APPEND is used to append data vertically in SAS in an already existing dataset.

Let us firstly initialize our library:

libname mylib '/home/u50789927/My_datasets';

Syntax:

PROC APPEND BASE = <base_data> DATA = <new_data_to_be_appended> <FORCE>;

RUN;

BASE_DATA : Data where output will be appended (If base data does not exist then a new data will be created)

DATA : Dataset whose rows need to be appended with the base dataset.

FORCE: To be used when new_data_to_be_appended has extra columns as compared to base dataset.

Case 1: When Base_data does not exist

In the following code we are appending an already existing data SASHELP.SHOES to a base data MYLIB.proc_append_output. Since it does not exist as of now thus a new data will be created.

PROC APPEND BASE  = MYLIB.PROC_Append_output DATA = SASHELP.SHOES;
RUN;

Case 2: When Base_data already exists

Task: Append SASHELP.SHOES again to the already existing dataset proc_append_output

PROC APPEND BASE  = MYLIB.PROC_Append_output DATA = SASHELP.SHOES;
RUN;

PROC CONTENTS DATA = MYLIB.PROC_Append_output;
RUN;

Note our resultant data has now 790 (395 * 2) rows.

Case 3: When New_data has a missing column.

Let us create a new base dataset to understand this named PROC_Append_output2

PROC APPEND BASE  = MYLIB.PROC_Append_output2 DATA = SASHELP.SHOES;
RUN;

Let us try to again append SASHELP.SHOES by dropping Sales column.

PROC APPEND BASE  = MYLIB.PROC_Append_output2 DATA = SASHELP.SHOES(Drop = Sales);
RUN;

SAS will create the dataset but will show a warning in the log! In the appended dataset, Sales will have missing values.

Case 4: When New_data has extra columns.

To understand this let us create a dummy date where Sales2 is a new column as follows:

DATA MYLIB.DUMMY_DATA;
SET SASHELP.SHOES;
Sales2 = Sales*2;
RUN;

Let us append this data in our already existing base data PROC_Append_output2

PROC APPEND BASE  = MYLIB.PROC_Append_output2 DATA = MYLIB.DUMMY_DATA;
RUN;

Appending in such a situation will lead to an error!

FORCE Statement:

In this situation FORCE statement comes to our rescue. It will not throw up the error, but a warning and will discard the extra column!

PROC APPEND BASE  = MYLIB.PROC_Append_output2 DATA = MYLIB.DUMMY_DATA FORCE;
RUN;