Allows the user to set the symbol that GAUSS uses when missing values are converted to ASCII and vice versa.
(literal or ^string (up to 8 letters)) if a string, and not surrounded by quotes, is forced to uppercase. This is the string to be printed for missing values. The default is ‘.’.
In the example below, you first create simulated data. The data represents the scores that a group of students received on a particular test and also the time that they took. For your calculations, you only want to consider data from students that completed the test in less than 80 minutes.
The code below replaces the scores from students that took more than 80 minutes with missing values. It uses the msym keyword to change the visual representation used for missing values from a ‘.’ to a ‘T’. Though, note that the underlying elements are still missing values, not character or string elements.
// Set seed for repeatable random numbers rndseed 543124; // Random integers with a mean of 70 and range of 20 to // represent time taken for test testTime = ceil(30 * rndu(10, 1)) + 60; // Random integers with a mean of 1000 and a standard // deviation of 10 score = ceil(10 * rndn(10, 1)) + 1000; // Maximum allowed time for test maxTime = 80; // Create a mask for times greater than maxTime mask = testTime .> maxTime; // Set scores to be missing values if testTime is greater // than maxTime mScores = missex(score, mask); // Set missing values to print as 'T' to represent that the // score was invalid because the student took too much time msym "T"; format /rd 4,0; print mScores;
The code above will return:
T 1010 997 1002 985 997 1007 995 T T
When converting ASCII to binary in
loadm and let statements, only the first character is significant. In other words,
will cause ‘H’ to be converted to missing on input.
This does not affect
writer(), which outputs data in binary format.
Note that msym is a keyword and not a variable being assigned to, so there is no equals sign between msym and the string that is being passed to it.