Scripting to Parse String with Delimiters

sophiaw

Hello JMP Community!

 

So I am trying to automate data analysis and my JMP script is very much hardcoded. I would like to make this more robust but I am having trouble with one part.

 

What my script is doing is it is going through all the rows of a column and parsing the values of the strings at each row. Since the data is collected in a string format such as: 0.002|0.003|0.004|0.005|.... etc. It is delimited with a '|'. My script makes 30 new columns (because that is how many values I am expecting in my delimited string) and then goes into nested for loops to parse all the data by the delimiter and put the values in order in the proper columns. What I just discovered is that maybe sometimes the string doesn't have all 30 values... It might only have 20. When that happens, JMP throws an error and exits my script. How do I get JMP to not care about that error and continue executing the script? (I added a Continue() function at the end which does execute the row that only has 20 values, but then it won't parse the data for the rest of the data table...)

 

This is my code:

 

 

dt << New Column("Year1", Numeric, Continuous, Format("Best",12))
    << New Column("Year2", Numeric, Continuous, Format("Best",12))
    << New Column("Year3", Numeric, Continuous, Format("Best",12))
    << New Column("Year4", Numeric, Continuous, Format("Best",12))
    << New Column("Year5", Numeric, Continuous, Format("Best",12))
    << New Column("Year6", Numeric, Continuous, Format("Best",12))
    << New Column("Year7", Numeric, Continuous, Format("Best",12))
    << New Column("Year8", Numeric, Continuous, Format("Best",12))
    << New Column("Year9", Numeric, Continuous, Format("Best",12))
    << New Column("Year10", Numeric, Continuous, Format("Best",12))
    << New Column("Year11", Numeric, Continuous, Format("Best",12))
    << New Column("Year12", Numeric, Continuous, Format("Best",12))
    << New Column("Year13", Numeric, Continuous, Format("Best",12))
    << New Column("Year14", Numeric, Continuous, Format("Best",12))
    << New Column("Year15", Numeric, Continuous, Format("Best",12))
    << New Column("Year16", Numeric, Continuous, Format("Best",12))
    << New Column("Year17", Numeric, Continuous, Format("Best",12))
    << New Column("Year18", Numeric, Continuous, Format("Best",12))
    << New Column("Year19", Numeric, Continuous, Format("Best",12))
    << New Column("Yearn20", Numeric, Continuous, Format("Best",12))
    << New Column("Year21", Numeric, Continuous, Format("Best",12))
    << New Column("Year22", Numeric, Continuous, Format("Best",12))
    << New Column("Year23", Numeric, Continuous, Format("Best",12))
    << New Column("Year24", Numeric, Continuous, Format("Best",12))
    << New Column("Year25", Numeric, Continuous, Format("Best",12))
    << New Column("Year26", Numeric, Continuous, Format("Best",12))
    << New Column("Year27", Numeric, Continuous, Format("Best",12))
    << New Column("Year28", Numeric, Continuous, Format("Best",12))
    << New Column("Year29", Numeric, Continuous, Format("Best",12))
    << New Column("Year30", Numeric, Continuous, Format("Best",12));
 
// Iterate through the rows of :Test_Result (where the delimited string is stored)
for (i = 1, i <= nrows(dt), i++,
    one_test_result_list = words(column(dt,"Test_Result"), "|");
  
    for (k = 1, k <= nitems(one_test_result_list), k++,
        // Put the values in the proper column at the proper row (i)
        dt:Year1 = num(one_test_result_list[1]);
        dt:Year2 = num(one_test_result_list[2]);
        dt:Year3 = num(one_test_result_list[3]);
        dt:Year4 = num(one_test_result_list[4]);
        dt:Year5 = num(one_test_result_list[5]);
        dt:Year6 = num(one_test_result_list[6]);
        dt:Year7 = num(one_test_result_list[7]);
        dt:Year8 = num(one_test_result_list[8]);
        dt:Year9 = num(one_test_result_list[9]);
        dt:Year10 = num(one_test_result_list[10]);
        dt:Year11 = num(one_test_result_list[11]);
        dt:Year12 = num(one_test_result_list[12]);
        dt:Year13 = num(one_test_result_list[13]);
        dt:Year14 = num(one_test_result_list[14]);
        dt:Year15 = num(one_test_result_list[15]);
        dt:Year16 = num(one_test_result_list[16]);
        dt:Year17 = num(one_test_result_list[17]);
        dt:Year18 = num(one_test_result_list[18]);
        dt:Year19 = num(one_test_result_list[19]);
        dt:Year20 = num(one_test_result_list[20]);
        dt:Year21 = num(one_test_result_list[21]);
        dt:Year22 = num(one_test_result_list[22]);
        dt:Year23 = num(one_test_result_list[23]);
        dt:Year24 = num(one_test_result_list[24]);
        dt:Year25 = num(one_test_result_list[25]);
        dt:Year26 = num(one_test_result_list[26]);
        dt:Year27 = num(one_test_result_list[27]);
        dt:Year28 = num(one_test_result_list[28]);
        dt:Year29 = num(one_test_result_list[29]);
        dt:Year30 = num(one_test_result_list[30]);
        Continue();
    );
);

 

 

 

 

Thank you for any help!!

ian_jmp

One way is to put the names of all your new columns into a list (for just three tests this would be 'testList = {"DDI1_TXP_0_SIU", "DDI1_TXP_1_SIU", "DDI1_TXP_2_SIU"};'). Then, the inner ('k') loop should be:

for (k = 1, k <= nitems(one_test_result_list), k++,
Column(dt, testList[) = num(one_test_result_list);
);

But note that this logic assumes that measured values are only 'missing to the right'(so in my example if DDI1_TXP_0_SIU was missing from the string and the other two were not, then the assignment would not be correct).