Subsetting a subset of Columns

[#1]

Hi All,

1. I have here a raw prober robot logs with continous by-time data per seconds of its movement executions a record of SiteX data (site0~7 this is basically a fixed hardware that touches a wafer), it records an X & Y and a BIN it touched down.

2. One row is one touch-down (of site0-7 ) on a wafer. However this raw format is hard to manipulate unless I do a subsetting and concatenating. See attached JMP raw data

Needed help:

1. I want to build a script that auto-subset the site0-7 columns and concatenate it like below.

2. Then I want to build a wafer map, that will track the sequence of touches the robot made on the wafer ove time continuous. See below table I want to make:

Anyone have idea? Thanks

Column 1	Column 2	Column 12	Column 13	Column 14	Column 15	Column 16	Column 17	Column 18	Column 19
8/14/2018	PM 11:59:38	Site0	X	29	Y	27	Pass	Bin	1
8/15/2018	AM 12:00:04	Site0	X	29	Y	25	Pass	Bin	1
8/15/2018	AM 12:00:30	Site0	X	30	Y	25	Pass	Bin	1
8/15/2018	AM 12:00:56	Site0	X	30	Y	27	Pass	Bin	1
8/15/2018	AM 12:01:22	Site0	X	34	Y	17	Pass	Bin	1
8/15/2018	AM 12:01:48	Site0	X	34	Y	18	Pass	Bin	1
8/14/2018	PM 11:59:38	Site1	X	28	Y	26	Pass	Bin	1
8/15/2018	AM 12:00:04	Site1	X	28	Y	24	Pass	Bin	1
8/15/2018	AM 12:00:30	Site1	X	29	Y	24	Pass	Bin	1
8/15/2018	AM 12:00:56	Site1	X	29	Y	26	Pass	Bin	1
8/15/2018	AM 12:01:22	Site1	X	33	Y	16	Pass	Bin	1
8/15/2018	AM 12:01:48	Site1	X	33	Y	17	Pass	Bin	1
8/14/2018	PM 11:59:38	Site2	X	27	Y	25	Pass	Bin	1
8/15/2018	AM 12:00:04	Site2	X	27	Y	23	Pass	Bin	1
8/15/2018	AM 12:00:30	Site2	X	28	Y	23	Pass	Bin	1
8/15/2018	AM 12:00:56	Site2	X	28	Y	25	Pass	Bin	1
8/15/2018	AM 12:01:22	Site2	X	32	Y	15	Pass	Bin	1
8/15/2018	AM 12:01:48	Site2	X	32	Y	16	Pass	Bin	1
8/14/2018	PM 11:59:38	Site3	X	26	Y	24	Pass	Bin	1
8/15/2018	AM 12:00:04	Site3	X	26	Y	22	Pass	Bin	1
8/15/2018	AM 12:00:30	Site3	X	27	Y	22	Pass	Bin	1
8/15/2018	AM 12:00:56	Site3	X	27	Y	24	Pass	Bin	1
8/15/2018	AM 12:01:22	Site3	X	31	Y	14	Pass	Bin	1
8/15/2018	AM 12:01:48	Site3	X	31	Y	15	Pass	Bin	1
8/14/2018	PM 11:59:38	Site4	X	25	Y	23	Pass	Bin	1
8/15/2018	AM 12:00:04	Site4	X	25	Y	21	Pass	Bin	1
8/15/2018	AM 12:00:30	Site4	X	26	Y	21	Pass	Bin	1
8/15/2018	AM 12:00:56	Site4	X	26	Y	23	Pass	Bin	1
8/15/2018	AM 12:01:22	Site4	X	30	Y	13	Pass	Bin	1
8/15/2018	AM 12:01:48	Site4	X	30	Y	14	Pass	Bin	1
8/14/2018	PM 11:59:38	Site5	X	24	Y	22	Pass	Bin	1
8/15/2018	AM 12:00:04	Site5	X	24	Y	20	Pass	Bin	1
8/15/2018	AM 12:00:30	Site5	X	25	Y	20	Pass	Bin	1
8/15/2018	AM 12:00:56	Site5	X	25	Y	22	Pass	Bin	1
8/15/2018	AM 12:01:22	Site5	X	29	Y	12	Pass	Bin	1
8/15/2018	AM 12:01:48	Site5	X	29	Y	13	Pass	Bin	1
8/14/2018	PM 11:59:38	Site6	X	23	Y	21	Pass	Bin	1
8/15/2018	AM 12:00:04	Site6	X	23	Y	19	Pass	Bin	1
8/15/2018	AM 12:00:30	Site6	X	24	Y	19	Pass	Bin	1
8/15/2018	AM 12:00:56	Site6	X	24	Y	21	Pass	Bin	1
8/15/2018	AM 12:01:22	Site6	X	28	Y	11	Pass	Bin	1
8/15/2018	AM 12:01:48	Site6	X	28	Y	12	Pass	Bin	1
8/14/2018	PM 11:59:38	Site7	X	22	Y	20	Pass	Bin	1
8/15/2018	AM 12:00:04	Site7	X	22	Y	18	Pass	Bin	1
8/15/2018	AM 12:00:30	Site7	X	23	Y	18	Pass	Bin	1
8/15/2018	AM 12:00:56	Site7	X	23	Y	20	Pass	Bin	1
8/15/2018	AM 12:01:22	Site7	X	27	Y	10	Pass	Bin	1
8/15/2018	AM 12:01:48	Site7	X	27	Y	11	Pass	Bin	1

[#2]

Attached is a script that uses the JMP Table Series Stack. From experience, people always state the table will alway have the same format, but that is not the case.  Note several of site columns were numeric and others were not.

Try the attached script on your prober table. Note that:

• the table name is hard-coded you might want to change that to Current Data Table(), or change it to an open.
• ns is set to 8 for the scenario of sites 0-7.  Change appropriately for other sites.
• run the script to just after line 24. You should verify the data stacked appropriately. Also you might want to sort it.

You should learn and understand each line of this script. You wil not be able to maintain it, if you do not understand it.   

[#3]

Hi gzmorgan0,

Thanks, I'm studying the codes,

I've deleted a Column 84 to 90 because I do not need this Col, only the site0-7 Col12-Col82 needs to be stacked due to its the usable same format. Sorry I didn't noticed at 1st.

I rerun the script again after deleting Col84-90 but got an error after:

idx = 1::ns*8;
cnames = (dt << get column names(character, "string"))[idx];  //error here

Error in jsl debugger

"Subscript Range in access or evaluation of 'Subscript' , (dt << get column names( character, "string" ))[/*###*/idx]"

I break the debugger to check at instance the raw data table, some Columns attribute were not changed to data type "character"

So I think because idx = 1::ns*8, index is 1~64 expects cnames count to be less than or equal to this index? 

cnames Col name holder  still stuck to this:

cnames={"Col12~Col90}

Hope you can still help me on this error.

[#4]

load your new table.

[#5]

Find attached

[#6]

I just ran the script on your new table and it ran perfectly. Which version of JMP are you using?

[#7]

JMP 12.0.1 (32bit)

[#8 · Accepted solution]

The cause of the error is that older versions of JMP were not consistent with the meessage << set. In JMP 14 it is much more consistent.

The error occured because the message for jmp 14 does not work in JMP 12.  Then the get column names for character columns was only 59 not 64 so idx (which is 1::64) threw and error.   I modified the script to use the current data table and used the older data type message. 

	column(dt,i) << set data type("Character")  //jmp 14
	column(dt,i) << data type("Character")   //jmp12

[#9]

Thank you, it works.