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
- :
- Is it possible to create an overlay plot representing confidence intervals?

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
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 22, 2015 3:12 PM
(1899 views)

I have lower bound and upper bound of the confidence interval for each estimate already available in my data, I just want to draw it using JMP. All the blogs I have found are regarding generating and drawing confidence intervals with JMP. Any help would be very welcome, thank you very much.

6 REPLIES

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 23, 2015 2:12 AM
(1736 views)

One way would be to use the 'Range Plot' option in Overlay Plot. To se an example, do 'File > New > New Script', cut and paste the code below, then do 'Edit > Run Script':

// Fudge some data

dt = Open**(**"$SAMPLE_DATA/Big Class.jmp"**)**;

UCIcol = dt << **NewColumn****(**"Weight UCI", Numeric, Continuous, Formula**(**:Weight + RandomNormal**(****10**, **1****)))**;

LCIcol = dt << **NewColumn****(**"Weight LCI", Numeric, Continuous, Formula**(**:Weight - RandomNormal**(****10**, **1****)))**;

// Use 'Range Plot' option in Overlay (add a 'Row()' variable to play the x role)

op = dt << **Overlay Plot****(**

X**(** Transform Column**(** "Row", Ordinal, Formula**(** Row**()** **)** **)** **)**,

Y**(** :Weight UCI, :Weight LCI **)**,

Range Plot**(** **1** **)**,

Show Points**(** **0** **)**,

SendToReport**(**

Dispatch**(** **{}**, "Overlay Plot Graph", FrameBox, **{**Frame Size**(** **688**, **243** **)}** **)**

**)**

**)**;

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 23, 2015 3:59 AM
(1736 views)

Hi Ian,

Thank you for your help. I am a new JMP user and so far have only been

using the menu options to generate graphs, manipulating all my data and

running models first in SAS. To create the overlay graph, I import by my

results dataset into JMP and select in menu bar "Graph" then "Overlay

plot". Is it possible to draw confidence interval using this menu, or

please could you explain to me how practically to run the code you sent me

earlier?

Thank you,

Myriam Alexander

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 23, 2015 5:44 AM
(1736 views)

Hello Myriam,

I'm sure you'll soon get used to JMP.

If you do want to run the code above, then the instructions to do that precede it. The first line to cut starts with '// Fudge . . . ' (the '//' actually denotes a comment line).

But, yes, it's easy to do it interactively. In the Overlay Plot launch dialog, assign the two columns with the upper and lower limits to the 'Y' role, and then the column you want to the 'X' role. When you hit OK, the report appears. Go to the little red triangle at the top of the report (in the 'Overlay Plot' outline node) to select the options you need. You should see 'Range Plot' as one of these.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 23, 2015 6:11 AM
(1736 views)

Hello Myriam,

You might find it easier using **Graph Builder**. Suppose you have a table with the following columns: X, Y, LCL and UCL. Bring up GB on this table. Drag X to the X axis. Highlight Y, LCL, UCL and drag to the Y axis. Click the icon for points, and then shift click the icon for vertical bars. Right-click inside the graph and select **Bar > Bar Style > Interval**. Then, under **Points** (left side) click on **Variables** and deselect **LCL** and **UCL**. Under **Bar** click on **Variables** and deselect **Y**. Should look like this:

If you run the code shown below inside a script window you'll get the final result. Uses JMP 11 btw.

New Table**(** "Untitled",

Add Rows**(** **10** **)**, New Column**(** "X", Numeric, Continuous, Format**(** "Best", **12** **)**,

Set Values**(** **[****1**, **2**, **3**, **4**, **5**, **6**, **7**, **8**, **9**, **10]** **)**

**)**,

New Column**(** "Y", Numeric, Continuous, Format**(** "Best", **12** **)**,

Set Values**(** **[****2**, **5**, **3**, **6**, **4**, **7**, **5**, **8**, **6**, **9]** **)**

**)**,

New Column**(** "LCL", Numeric, Continuous, Format**(** "Best", **12** **)**,

Set Values**(** **[****1**, **3**, **2**, **2**, **2**, **5**, **3**, **7**, **5**, **7]** **)**

**)**,

New Column**(** "UCL", Numeric, Continuous, Format**(** "Best", **12** **)**,

Set Values**(** **[****3**, **8**, **5**, **8**, **6**, **9**, **8**, **9**, **7**, **10]** **)**

**)**

**)**;

Graph Builder**(**

Show Control Panel**(** **0** **)**,

Variables**(** X**(** :X **)**, Y**(** :Y **)**, Y**(** :LCL, Position**(** **1** **)** **)**, Y**(** :UCL, Position**(** **1** **)** **)**

**)**,

Elements**(**

Points**(** X, Y**(** **1** **)**, Legend**(** **4** **)**, Jitter**(** **1** **)** **)**,

Bar**(** X, Y**(** **2** **)**, Y**(** **3** **)**, Legend**(** **5** **)**, Bar Style**(** "Interval" **)**,

Summary Statistic**(** "Mean" **)**

**)**

**)**

**)**;

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 24, 2015 3:19 AM
(1736 views)

Thank you very much for this step by step explanation, this answers my

question.

Best wishes,

Myriam Alexander

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Nov 23, 2015 7:16 AM
(1736 views)

Indeed, as Peter says Graph Builder may be a better choice - Possibly I over interpreted the 'overlay plot' in the title of the thread.

As a relative newcomer to JMP, you should be aware that there are almost always many ways to get things done - That's one of its charms . . .