>
> From: Steve Hodge <steve@nmr.mgh.harvard.edu>
> Date: 2006/01/06 Fri PM 05:21:05 EST
> To: jmp-l@lists.biostat.wustl.edu
> Subject: Re: [jmp-l] exclude and hide row vs exclude and hide cell
>
> Ron,
>
> If I understand correctly, you'd like to set "Cell States" (hide,
> exclude, label, etc) to particular entries in the data table. Sort of
> like an extension of Row States.
That's exactly what I want to do.
>
> I don't know if this facility exists in JMP.
For a $600 dollar upgrade, why not? Maybe this should be a part of JMP feature
request. I will save my $600.00 and get to work on the script. I may have to
come back to this site from time to time for advice.
>
> Maybe you could devise a script that would make a first-pass browse
> through the data table and create a table (a matrix) of the same
> dimensions as the data table. The values could be 0=good and 1=outlier,
> corresponding to whether the value at the same coordinates of the real
> data is an outlier or not. Then run the analyses with the scripting
> language, but set the row states immediately before running the test. The
> row states would be set by looking up entries in the "outlier matrix" that
> you made. After running the test, re-set the row states before moving to
> the next column.
It sounds like what you are proposing is the following:
set up a matrix of good vs bad data.
repeat until all columns are plotted in Fit y by x
get i'th column (data is numeric and continuous) and set row state. Row
state is set by the matrix for that column
run Fit Y by x on i'th column.
get next column i'th = i'th + 1
reset row state
end repeat
>
> Would this even work?? Ack!
This would work, but it will take some time. I know a little about jmp
scripting so I don't mind working on this. Is anyone else interested in a
script that would preserve the data table and allow plots that don't exclude
the entire row across every column?
One of the requirements would be to have all of my plots within the same dialog
box. I don't think that I can deal with 500-600 individual dialog boxes.
I think that a distinction should be made between outlyers and missing data.
thx
RLC
>
> Steve
>
>
>
> On Fri, 6 Jan 2006, Ronald Cross wrote:
>
> > Hi all, I am using JMP 4.0.4 for mac running a G4 powerbook with
> > 1GByte of RAM with a 1GHz power pc processor on OS X 10.4.3.
> > However, this version of jmp runs under OS 9 (classic).
> >
> > I am looking for suggestions on how to manage data, particularly
> > outlyers. I generally have hundreds (500-700) of columns of data to
> > look at for a given set of nominal condition.
> >
> > Currently, if I want to exclude data at a given row and column, I
> > have to do a list check and eliminate the outlyer-which then becomes
> > missing data: when the data is missing, I am not able to retrieve it.
> >
> > The question that I have is how can exclude and hide a cell at a
> > given specified row and column without losing the data? I don't want
> > to exclude and hide the entire row because it effects every column in
> > the data table, and these columns may be independent from one
> > another, i.e. a bad data point at row x and col y doesn't necessarily
> > mean a bad data point at row x and col z.
> >
> > Is this fixed in the latest version of JMP for Macs?
> >
> > thx
> > RLC
> >
> >
> >
> >
> >
> >
>
>
|