What inspired this wish list request? Survey data frequently contains missing values. The ideal data set either uses a period (.) or a distinct missing value (e.g. -9999) with a Missing Value Code column property to designate a missing value. Now, let's assume that a column with a formula column property is created from existing columns containing distinct missing value data (e.g. NewCol = sum(Col1, Col2) ). A situation may arise where the distinct missing value (used by the database creator) is discovered later (by the data analyst) and then the Missing Value Code column property is added. Currently, JMP does not automatically Rerun Formulas to account for the missing value code property. So, in this case, even though Col1 and Col2 will now treat -9999 as missing, the formula column (NewCol) will still evaluate -9999 as -9999 if it encounters it inCol1 or Col2.
What is the improvement you would like to see?
I understand that it is possible to go to the data table and select Rerun Formulas from the drop down triangle, but the "wish" is that JMP reruns formulas automatically when a missing value code (or any other column property that impacts the value of the column) is changed. This will prevent an unintended miscalculation of the formula column that the user might assume is updated along with the missing variable columns.
Why is this idea important? It seems an obvious thing to do when you consider the circumstances. Without manually performing the Rerun Formulas option, the user may accidentally report incorrect formula values that they assumed were fixed when missing value code were added. It's a quirky situation, but that's what makes it a potential problem for the unsuspecting analyst.
JMP 18 User