A histogram provides an overview of the distribution of data in the file. Often, the distribution details can provide the analyst/internal auditor with ideas on which types of further analysis may be desirable.
Output from the histogram is written both as data which can be imported into other programs such as MS-Excel, as well as output in html format and a chart for ease of viewing.
The syntax for the analytical specifications is fairly straightforward. Required are the starting value (i.e. lowest value for which histogram values are to be obtained), the number of "bins" to be used (maximum is 100), and the width of each bin (all bin widths are uniform).
Output from the procedure is both tabular and visual, as presented in HTML, in order to streamline the analysis process for the internal auditor.
Syntax (Batch Script processing)
proc histo data=DATAFILE start=START width=WIDTH n=N;
var VARIABLENAME;
out = OUTPUTRESULT;
Parameters Used
There are six primary parameters for the Histogram chart procedure:
DATAFILE - the datafile to be analyzed
VARIABLENAME - the numeric variable to be analyzed
OUTPUTRESULT - name of the file to store the results of the analysis
START - the lowest number for the range of data to collect histogram results
WIDTH - the width of each bin in the histogram
N - number of bins (maximum of 100)
Example Script
*;
* histo1.ezs;
* obtain a histogram chart to display the distribution of data;
* besides the tabular output, also creates an HTML viewable table;
*;
libname test 'c:/ezs/tab';
libname drug 'c:/ezs/tab/tested';
proc histo data=test.maud2002 chart vertical start=.1 width=.5 n=30;
out=drug.h;
var QUAN1;
run;
Example Output
Two output files are created, the text file and a chart file. The text file is a tab separated file.Analysts interested in Proc Histo may also be interested in Proc Means and Proc Univariate, which are also available at no charge.