My assumption is that since you are not specifically scoping your variables (:), that you are running into issues when trying to run in the 2 different environments. The other item I observed is that your method is going to result in a very slow piece of code. Given that, I have rewritten your script, using some methods that I believe will work in both environments, and will also run much faster. The script below run against the Semiconductor Capability data table that is deistributed with JMP. But if you just strip out the reference to opening that table, along with the statement following the Open of the data table and replace them with your
dt = current data table();
the code should work on your data.
I have annotated the code below, so I think you should be able to follow it. Also, I VERY STRONGLY SUGGEST that you read the Scripting Guide distributed with your install of JMP.
Help==>Books==>Scripting Guide
It will tell you all about what those crazy colons are all about
Names Default To Here( 1 );
dt = Open( "$SAMPLE_DATA/semiconductor capability.jmp" );
dt << delete columns( {2, 3, 4} );
// Get the numeric column names as strings
colNames = dt << get column names( Numeric, string );
// Create the new calculated columns
// They will be populated later
dt << New Column( "Num Tests Failing", numeric );
dt << New Column( "Failing Tests", character );
// Create a new column for each column in the colNames list
// and populate the colNames name, if it does NOT pass
For( i = 1, i <= N Items( colNames ), i++,
// Create the new colum
dt << New Column( "PassFail " || Char( i ), character );
// Get the Spec Limits for the first column
specLimits = Column( dt, colNames[i] ) << get property( "Spec Limits" );
// If no Spec Limits are found, do not process
If( Is Empty( specLimits ) == 0,
// Get the USL and LSL and if not present set to missing
LSL = Try( specLimits["LSL"], . );
USL = Try( specLimits["USL"], . );
// If the LSL is not missing process
If( Is Missing( LSL ) == 0,
// Find all rows where the colNames column is less than
// the LSL
dt << select where( As Column( dt, colNames[i] ) < LSL );
// If any rows are selected, then process
If( N Rows( dt << get selected rows ) > 0,
// Set the value in the PassFail column to the name of
// the colNames being processed. This will allow for
// the counting and the concatenating for the new
// calculation columns
Column( dt, "PassFail " || Char( i ) )[dt << get selected rows] = colNames[i]
|| "; "
);
);
// See above comments for LSL
If( Is Missing( USL ) == 0,
dt << select where( As Column( dt, colNames[i] ) > USL );
If( N Rows( dt << get selected rows ) > 0,
Column( dt, "PassFail " || Char( i ) )[dt << get selected rows] = colNames[i]
|| "; "
);
);
);
);
// Clean up and selected columns
dt << clear select;
// Create the command string required for the formula for the
// Failing Tests, which is a concatenated string of all of the
// PassFail columns.
theExpr = "dt:Failing Tests << set formula(Concat(:PassFail 1";
For( i = 2, i <= N Items( colNames ), i++,
theExpr = theExpr || ",:PassFail " || Char( i )
);
theExpr = theExpr || "))";
// Execute the command string
Eval( Parse( theExpr ) );
// Delete the formula, which will remove the formula, but not
// the calculated values
dt:Failing Tests << delete formula;
// Create the command string for the Num Tests Failing column
// The count is the sum of the "True" Boolean comparisions.
// That is, if PassFail 1 != "", it will be "True" and therefore
// will have the value of 1. If the cell is blank, it will have a
// value of 0. The sum of all of these comparisons will be equal to
// the number of non passing parameters(tests)
theExpr = "dt:Num Tests Failing << set formula(sum(:PassFail 1 != \!"\!"";
For( i = 2, i <= N Items( colNames ), i++,
theExpr = theExpr || ",:PassFail " || Char( I ) || " != \!"\!""
);
theExpr = theExpr || "));";
Eval( Parse( theExpr ) );
dt:Num Tests Failing << delete formula;
// Delete all of the PassFail columns, just to clean up the data table
dt << delete columns(index(n cols(dt)-N items(colNames)+1,ncols(dt)));
Jim
