Hi,   I accidentally started a logistic fit that is/will take forever to run.  The progress window is shown with cancel, but clicking it does nothing.  I would really like to avoid force-quiting the program because of some unsaved work.   Are there any undocumented ways of stopping the fit? JMP 12 Mac OS. Andy ... View more

Hi,   I am working with some MySQLdatabase integrating with JMP 12. (No I cannot afford to upgrade). I am seeing some data corruption when I import a table. Using MySQLWorkbench to view the data.  Notice the first 2 rows id = A01B and its associated title and compare it to the JMP table.  The id value is trunctated and title is missing.  In row 2, A01C, the title value is corrupt.   Is this a known issue and is there a work around? ... View more

Hi,   My goal is a script that loads all the csv files in a directory and then does name clean up.  Previously I had developed a script that worked on a single file and I am looking to expand it to batch loading.  I found a script on the board combining txt tables.jsl and added in the code so it works on the Mac and inserted my settings for the txt import.  That seems to work.   I am now working on the column clean up and this is the script.  It has trouble finding the "Column 3".  Up til then it executes.  Now the odd thing rhese lines were copied from my original script. If I highlight the line with the column 3 rename in the new script and hit run (so only that line executes) I get an error.  If I highlight the same line in my original script and execute it works as intended.  I have literally copied and pasted (command-C and command-V) between the scripts.   What am I missing? And yes Column 3 exists.   Andy dt=current data table(); dt<<delete columns("Column 1"); dt<<delete columns("Column 2"); dt<<set table variable("Thickness",0.010); dt<<set table variable("RotLen",0.16); dt<<new column("meter",continuous,set formula(Row() * :RotLen)); //here is where it barfs. name("Column 3")<<set name("Roller"); name("Column 4")<<set name("RawCen"); name("Column 5")<<set name("RawFar"); name("Column 6")<<set name("RawNear"); name("Column 7")<<set name("Center"); name("Column 8")<<set name("Far"); name("Column 9")<<set name("Near"); :Center<<set formula(:RawCen - :Thickness -:roller*0.50); :Far<<set formula(:RawFar-:Thickness-:roller*0.25); :Near<<set formula(:RawNear-:Thickness-:roller*0.75); dt<<go to(:Far); dt << Move Selected Columns( To first ); dt<<go to(:Near); dt << Move Selected Columns( To first ); dt<<go to(:Center); dt << Move Selected Columns( To first ); dt<<go to(:meter); dt << Move Selected Columns( To first ); ... View more

Hi,   I am working on script and ran into problem/bug.  I would to validate this or at least understand what is happening. Here is what I am trying to do: I have a time series of data for a bunch of samples and I want to generate a column that compares the measured value to that when the sample was first meaured. Basically a (Y0-Y)/Y0.   Do do that I generate a subset where the first data reading for each sample (which may not be count =1).  I then get the important parameters by doing a table update to the subset.  Changes some names for clarity purposes and Then update the original data table with the intial reading data such that I have a column of Y and seperate column of Y0.   The last step causes a problem. The subset table has a column names Cycle_num which represents the minimum value on a per sample basis.  The orginal also has one with that name but includes all the different point.  The last table update matches on sample id and specifies the columns to add.  The list specifically does NOT include cycle_num. Nor is cycle_num used as a match column.  Upon table updating  the data in cycle_num in the original table has been changed to the values from the subtable.  The work around is to delete the cycle_num column from the subtable before using it in the last update step.   Has anyone seen a bug like this or are all columns sharing names always used in matches and updated even if not specifically called to?   Andy   //Get initial Data note: first run may not be at 1 so use min dt1=dt<<Summary( Group( :device_id, :sweep_direction,:temporal_point_type ), Min( :cycle_num ), Freq( "None" ), Weight( "None" ), statistics column name format( "column" ), Link to original data table( 0 ) ); //get rid of mid values dt1<<Select Where(:temporal_point_type == "mid"); dt1<<delete rows; //get important initial data dt1 << Update( With( Data Table( dt) ), Match Columns( :device_id = :device_id, :sweep_direction = :sweep_direction, :temporal_point_type = :temporal_point_type, :cycle_num = :cycle_num ), Add Columns from Update table( :tvis_bright, :tvis_dark, :l_star, :a_star, :b_star, :Name( "ChargeTransfer [mC/cm2]" ) ) ); // Do some clean up to prevent name collision and bring into base table Data Table (DT1); :tvis_bright<<Set Name("TVB0"); :Name( "ChargeTransfer [mC/cm2]" ) <<Set Name("CT0"); :tvis_dark<<Set Name("TVD0"); :l_star<<Set Name("LS0"); :a_star<<Set Name("AS0"); :b_star<<Set Name("BS0"); dt1<<delete columns(:cycle_num); //added as work around solution dt<< Update( With(Data Table(Dt1)), Match Columns( :device_id = :device_id, :sweep_direction = :sweep_direction ), Add Columns from Update table(:TVB0,:TVD0,:CT0,:LS0,:AS0,:BS0 ), ); ... View more

Hi,   Often I get data with full time date stamps, really just the seconds since jan 1 1904.  What I would like is a simple way to turn this into a date.  Lump everything from midnight through 11:59PM.   Yes I can write a formula to do this but it would easier if A. there was a Time format that would truncate  this 01/26/2018 10:47 PM  and 01/26/2018 4:13 AM to 01/26/2018 and model it as such.  I can display it as a date, but it is modeled so they would be the same value.   Failing that   B. An option from the equation builder to return the date.  Yes I know I can write with a formula, but I think more than myself is doing this.   Andy ... View more

More a thread to think about how to best describe variability than a pure JMP question.   The interest comes with a background from the semiconductor industry, but has applictions elsewhere.  What metrics should be used to describe the variability over a part, a wafer in this case.   The first answer is always variance/standard deviation, but here is a bit more background.  When we judge the results of a process we typically take lots  of measurements across the wafer.  The locations of those points are NOT random.  Typically either a series of points across a diameter or a concentric series of rings are taken.  There are often real limitations to where measurements can be made and there is really no options to change them.   The second complication is that for the most part the processes have very systematic trends across the wafer and a random noise factor on top of that.  Patterns are often center high/ edge low or sometimes more complex but predictable.   Currently two metrics are used: standard deviation or Max/Min (range).  I have looked at results for many decades now and find either metric wanting.  Typically a standard deviation predicts a much larger range of results than is actually seen and the Max-Min techniqueactually doesn't even capture the points measured.  Side Note the metric is Max-Min/2Average, but if the average is not in the center of the range, then this metric will not include the most extreme values.   The central limit theorom says the non-random point and systematic variations is okay in calculating the means, but it really doesn't address the variability or range of measurements.   Should I just stop worying about this or is there a better solution?  One reason for the concern is that in doing optimaztion calculations the two common metrics give two different set of optimal conditions.   Thoughts? ... View more