JMP Connections

APPENDIX BJMP^® JSL Time Constant Learning Curve Script

// open the database and initialize the table with columns1
   from the die_cut_stamp_oee view
dt = Open Database("DSN=Heavy␣Metal␣Metrics;UID=jwubbel;PWD=2
   geu33b;SERVER=localhost;DATABASE=micro_data_warehouse;PORT
   =3306;", "SELECT␣*␣FROM␣DIE_CUT_STAMP_OEE␣ORDER␣BY␣Date␣
   ASC", "PLANT␣29␣OEE␣SUPER␣STAMPER␣CONTROL␣CHART");
3
// add in the column for sequencing required by the Time4
   Constant Learning Curve algorithm
dc_seq = dt << Add Multiple Columns( "seq", 1,5
    "Before␣First", numeric, continuous, best, 3 );
6
// initial the seq column with row numbers7
For Each Row(seq=Row());8
9
// count total number of rows in the current table10
iRowCnt = N Row(dt);11
12
// add new columns13
14
// create new columns for facilitating the newton-gaussian15
   non-linear regression calculations
dc_nc = dt<<NewColumn("Errors",numeric,best);16
dc_nc << Set Modeling Type("Continuous");17
18
dc_lp = dt<<NewColumn("Estimated␣OEE",numeric,best);19
dc_lp << Set Modeling Type("Continuous");20
21
// set up script to recalculate the learning curve starting22
   values for the "Estimated OEE" parameters
dc_lc = dt<<NewColumn("x-formula",numeric,best);23
dc_lc << Set Modeling Type("Continuous");24
25
// set the Time Constant Learning Curve formula for the26
   column "Estimated OEE"
dc_lc << Add Column Properties(27
        Formula(28
                Parameter(29
                        {Yc = 57.6, Yf = 10, t =30
                20},31
                         Yc + Yf *(1 - Exp(-seq/t))32
                )33
        )34
);35
36
// Define script to run ...