Subscribe Bookmark RSS Feed

Scatterplot x-axis values ?

Drharsha

Occasional Contributor

Joined:

May 2, 2017

I am ploting a stifness matrix plot with respect ot temperature from 300 to 1200 kelvin but when i am ploting the temperature values. but if you can see the x-axis the temperature values are jumbled. I actually want the m in order i.e 300 kel 400 kel 500 kel etc but it is jumbled i.e 400 kel 300 kel 500 kel etc. plz tell me what can i do know.

8 REPLIES
txnelson

Super User

Joined:

Jun 22, 2012

I appears that your y xxis values are a character string, and not numeric.  Therefore the values are ordered alphabetically.

Change it to a numeric and it should work.

 

or possibly you have the column property "Row Order Levels" selected.

Jim
Drharsha

Occasional Contributor

Joined:

May 2, 2017

i am sending the picture of the input dat

Screen Shot 2017-05-03 at 12.29.07 AM.png

Dan_Obermiller

Joined:

Apr 3, 2013

Actually, it looks like you have several numeric columns that are all entered as text strings. You can tell because they are left-justified within the column. Numbers are always right-justified. I would recommend doing some data cleanup to get everything in the proper format before doing any analysis. Remember that although JMP looks like a spreadsheet, it is not. The first step in any good analysis is to structure the data properly.

Dan Obermiller
Drharsha

Occasional Contributor

Joined:

May 2, 2017

how to make these numeric to left justified, is there any short cut for it. I am new to this software so plz dont mind in giving an answer.

txnelson

Super User

Joined:

Jun 22, 2012

OK.....there really isn't a format that will left justify a numeric field in a data table.  However, there is a way that after you convert your values to numeric, that you can left justify them in the data table.  Now, this is not really a published method for doing this, but if you really need to have numeric values left justified it does work.

 

One of the column properties that you can apply to your columns is called "Value Labels".  They allow you to specify a different value in the table, than is actually recorded.  A typical example is when you have a column called "Gender", where you have recorded 1 for Males and 2 for Females.  So by using Value Labels, you can specify that rather than displaying the numeric values, use the character values.  Because they are now displayed as characters, they will be left justified.   If you are following this, you are probably thinking that what I am specifying, is that I am going to tell you that you will need to add a value label for each numeric value found in your data.  Well, you could do that, except that isn't practical.  But there is a trick.  It turns out, that once a single Value Label is specified, JMP assumes that the column is to be displayed as a character string.  And because of that, it left justifies the column in the data table.

So, all you need to do, is to add one Value Label to a column, and I suggest that you specify a value that isn't even a value that appears in the real data.  

Below I have an example using the distributed sample data table, "Big Class"

Here are the Value Labels I set for the numeric columns Height and Weight

vl 1.PNGvl2.PNG

Here is the data table, as it is displayed after setting the Value Labels

vl3.PNG

And here is a graphical display showing that JMP still displays the values in numeric order, not alphabetical

vl4.PNG

 

The last thing, is that all you have to do, is to set the Value Labels for just one of your columns, and then you can use Standardize Attributes to copy the Value Labels to all of the columns you need to have left justified

 

Jim
M_Anderson

Staff

Joined:

Nov 21, 2014

Adding to what @txnelson said, I'd change the temperature column to an Ordinal Numeric column and put K in as a "Units" column property (Kelvin is "K" not "kel", btw). That also gives you the option to switch the column to Continuous Numeric if you want to do some regression or something.

Best,

M
ian_jmp

Staff

Joined:

Jun 23, 2011

Maybe stating the obvious, but you can make the changes Mike suggests by double clicking on the heading of a column to bring up the Column Info dialog.

markbailey

Staff

Joined:

Jun 23, 2011

Also, since you need to make the same changes to more than one column, select all of the affected columns and then select Cols > Standardize Attributes. You have to enable each attribute in turn for safety's sake and then make the change. All of the columns will be updated when you click OK.

Learn it once, use it forever!