Answer
Here are some examples of how you would create a series plot, bar chart, pie chart, and band plot in a SAS Code node in Model Studio, using whatever data that you create or reference in the node in the &dm_lib library (just replace the name exampData with your name, and the actual data with the data you want to use).
/* SAS code */
data &dm_lib..exampData;
a = "a"; b = 2; c=1; d=3; x=3;
output;
a = "b"; b = 5; c=4; d=9; x=5;
output;
a = "c"; b = 4; c=2; d=5; x=10;
output;
run;
/* Series Plot */
%dmcas_report(dataset=exampData, reportType=SeriesPlot, x=a, y=b, description=%nrbquote(Series Plot), yref=3);
/* Bar Chart */
%dmcas_report(dataset=exampData, reportType=BarChart, category=a, response=b,description=%nrbquote(Bar Chart));
/* Bar Chart - sorted X-axis */
%dmcas_report(dataset=exampData, reportType=BarChart, category=a, response=b, sortBy=b, description=%nrbquote(Bar Chart - Sorted));
/* Pie Chart */
%dmcas_report(dataset=exampData, reportType=PieChart, category=a, response=b, description=%nrbquote(Segment Plot));
/* Band Plot */
%dmcas_report(dataset=exampData, reportType=BandPlot, x=x, y=b, limitLower=c, limitUpper=d, description=%nrbquote(Band Plot));
Here’s an example for creating a scatter plot of Residuals by Predicted in a SAS Code node following a Supervised Learning node for an interval target:
/* SAS code */
data &dm_lib..samp;
set &dm_data(obs=200);
Residual = Target - P_Target; /* substitute "Target" with the name of your target variable */
run;
%dmcas_report(dataset=samp, reportType=ScatterPlot, x=P_Target, y=Residual, description=%nrbquote(Scatter Plot), yref=0);
These code examples are also available from this GitHub repository.
