turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- JMP User Community
- :
- Discussions
- :
- Discussions
- :
- How to change X axis values (a numeric column) by ...

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Feb 18, 2009 11:19 AM
(1301 views)

Hi, I have the following 3 columns:

Col_1 Col_2 Col_3

00Qtr1 1 1.4

00Qtr2 2 1.5

00Qtr3 3 1.46

00Qtr4 4 1.6

01Qtr1 5 1.54

01Qtr2 6 1.58

01Qtr3 7 1.69

01Qtr4 8 1.6

02Qtr1 9 1.62

... .... ...

08Qtr4 28 2.48

I have tried Fit Y(Col_3) by X(Col_2), because both are numeric data type, for linear regression.

In the picture box or plot, I want to repley X axis values {1, 2, 3, 4, ....} by

the values of Col_1, i.e. {00Qtr1, 00Qtr2, 00Qtr3, ...}

So how to revise the following JSL code to do that?

Thanks,

Fred

-----------------------------------------------------------------------------------------------

biv = Bivariate(

Y( : Col_3 ),

X( : Col_2 ),

Fit Line( {Line Color( "Red" )} ),

SendToReport(

Dispatch(

{},

"1",

ScaleBox,

{Scale( Linear ), Format( "Best" ), Min( 1 ), Max( 28 ),

Inc( 4 )}

),

Dispatch(

{},

"2",

ScaleBox,

{Scale( Linear ), Format( "Numeric" ), Min( 0 ),

Max( 2.5 ), Inc( .5 )}

),

Dispatch( {}, "Bivar Plot", FrameBox, Marker Size( 2 ) )

)

);

Col_1 Col_2 Col_3

00Qtr1 1 1.4

00Qtr2 2 1.5

00Qtr3 3 1.46

00Qtr4 4 1.6

01Qtr1 5 1.54

01Qtr2 6 1.58

01Qtr3 7 1.69

01Qtr4 8 1.6

02Qtr1 9 1.62

... .... ...

08Qtr4 28 2.48

I have tried Fit Y(Col_3) by X(Col_2), because both are numeric data type, for linear regression.

In the picture box or plot, I want to repley X axis values {1, 2, 3, 4, ....} by

the values of Col_1, i.e. {00Qtr1, 00Qtr2, 00Qtr3, ...}

So how to revise the following JSL code to do that?

Thanks,

Fred

-----------------------------------------------------------------------------------------------

biv = Bivariate(

Y( : Col_3 ),

X( : Col_2 ),

Fit Line( {Line Color( "Red" )} ),

SendToReport(

Dispatch(

{},

"1",

ScaleBox,

{Scale( Linear ), Format( "Best" ), Min( 1 ), Max( 28 ),

Inc( 4 )}

),

Dispatch(

{},

"2",

ScaleBox,

{Scale( Linear ), Format( "Numeric" ), Min( 0 ),

Max( 2.5 ), Inc( .5 )}

),

Dispatch( {}, "Bivar Plot", FrameBox, Marker Size( 2 ) )

)

);

1 REPLY

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Feb 19, 2009 6:01 AM
(1268 views)

I think only Scatterplot Matrix and Graph Builder can treat a nominal column like a numeric one. They interpret the nominal levels as 0..n-1 which is important if you combine them with other elements.

For instance, if you want to show a fitted regression, you could

1) change Col_2 to contain 0..27

2) use one of the above platforms to plot Col_3 vs Col_1

3) use bivariate to fit Col_3 vs Col_2

4) toggle the bivariate Show Points option so only the line is present

5) right-click in the bivariate plot and choose Edit > Copy Frame Contents

6) right-click in the other plot and choose Edit > Paste Frame Contents

For instance, if you want to show a fitted regression, you could

1) change Col_2 to contain 0..27

2) use one of the above platforms to plot Col_3 vs Col_1

3) use bivariate to fit Col_3 vs Col_2

4) toggle the bivariate Show Points option so only the line is present

5) right-click in the bivariate plot and choose Edit > Copy Frame Contents

6) right-click in the other plot and choose Edit > Paste Frame Contents