Overriding the JMP Excel Import Wizard Column Type

Question

Hi,

Is there a way to force JMP to import a column as a character column when it thinks it's numeric?  I have a column with serial numbers that contains numbers and letters.  But for some reason JMP thinks it's a numeric column so it always deletes the values in the column?  I've tried changing all the values in the wizard but it always thinks the column is numeric.  Is there a way to force it import as is?  Thanks

DL723 · Answer

I found that if i just throw an open file command in the jmp script that seems to work.  It sets all the columns in that sheet as character.  Not ideal since I have to throw some code to close the unecessary sheets and change column types of all the other columns but at least i can automate it.  Would be interesting to know why the wizard thinks that column is numeric though.

DL723 · Answer

Hi Brian,

Thanks for the feedback.  Not sure if you meant send an email to tech support or just post on here..but here's a sample data set.  I basically have the wizard setup to pull the second table starting with the Serial header.  I did notice it works with bigger data sets which supports the 100+.  But unfortnately not all tables, that i'm trying to automate pulling, have large sets.

I also tried pulling the whole sheet (with that summary on top) but it still seems to treat column 1 as numeric.

Lot Number123456   XXYYZZ  -3.9714.33Average -3.670.29Standard Deviation     SerialXXYYZZ 811699-A 2.9914.28 811699-B 3.5014.19 811699-C 2.9514.55 811699-D 3.2014.31 811699-E 2.9214.52 811699-F 3.2414.58 811699-G 3.0514.74 811699-H 3.1914.80 827662-A 3.2814.12 827662-B 3.2514.36 827662-C 3.4113.80 827662-D 3.4713.90 827662-E 3.3914.27 827662-F 3.4813.92 827662-G 3.3714.02 827662-H 3.2714.89 848230-A 2.8414.50 848230-B 3.0614.45 848230-C 2.9614.23 848230-D 3.0214.19 848230-E 3.1814.10 848230-F 21.18  848230-G 3.4714.24 848230-H 3.5714.68

briancorcoran · Answer

If you have more than 100 rows of data you can try selecting the checkbox "Show all Rows" to make sure the type detection mechanism is looking at all of the data.  Otherwise, there is no way to coerce a character column without putting some character data in the end or something like that.  If you can share the data with Tech Support, I would encourage you to do that so we can figure out why your values are not being treated as character.

Brian Corcoran

JMP Development