Student Edition Needs Curve Fitting Coefficients

[jjproctor]

Students need to be able to analyze data and perform curve fitting for their assignments.  With JMP licenses being out of reach for them (cost is greater than $1,000), this functionality needs to be made available to them.  This functionality comes for free using numpy, Excel, or other tools and should be free in JMP if we wish for students to learn the platform in hopes of them using it during their professional careers.

[julian]

Hi @jjproctor

Can you expand on what kinds of curves you wish students to fit with JMP Student Edition? In JMP Student Edition, both Analyze > Fit Y by X, and Analyze > Fit Model allow for curve fitting by adding polynomial terms to the models. This is done most easily in Analyze > Fit Y by X using the Red Triangle option of Fit Polynomial, where you may choose models with up to a 6th degree polynomial. The option of Fit Special from the same Red Triangle also allows for curve fitting by way of transformations to the Y or X. If you are looking for nonlinear models (e.g., sigmoid curve, exponential growth and decay, or pharmacokinetic models), that is indeed functionality that is only available in standard versions of jmp at this time. That said, for students the price is not $1000+. Students may get their own annual license for standard JMP from onthehub.com for less than the cost of an inexpensive textbook, or through an unlimited university licence available at many schools.

I hope this helps!

Julian

[jjproctor]

Thank you for the quick response. Unfortunately, I do not see the options
you refer to. Please see attached screen shot.
I don't know how to proceed.

[image: image.png]

[jjproctor]

For Fit Y By X: I see blue triangles, but the red triangle does not have the features listed .Here is the picture:

[jjproctor]

For Fit Model: I see blue triangles, but the red triangle does not have the features listed .Here is the picture:

[julian]

Hi @jjproctor,

I think a few different things are happening here that are preventing you from seeing the options you want. I highly suggest watching some of our getting started videos because these actions are shown there in a more visual way. Also, under the Help menu there is a Discovering JMP Student Edition book that explains the basic use of JMP. That said, let's get you those analysis you are looking for:

We need to address a few things: First, the modeling type for both of your columns is set incorrectly for the kind of analysis you wish to perform. This is important because JMP is intelligent about the kinds of analyses it offers based on what modeling type the data in the table is set to be.  If you look in the columns list on the left-hand side of the screen, you will see a red histogram icon next to each column name. That indicates that JMP is treating those columns as collections of nominal observations. In other words, that each value in that column is a category, rather than a measurement on a scale variable. When you click one of those red histograms you can change the modeling type. What you want here is "Continuous."  Give that a try, but I suspect what might have happened is that when you imported your data there were words or letters in some of the columns, and so JMP set not just the modeling type of Nominal, but the Data type of the column to Character. If that happened, you will need to right-click the column header and select Column Info. Then, under Data type, change the option to Numeric, and under Modeling Type, change the option to Continuous. When you click okay, you'll now see blue triangles next to the column name for each of your variables. That means JMP will now treat those measurements as actual numbers. 

When you launch Fit Y by X, enter your columns as before then click OK. Once you are into the platform, click the Red Triangle at the top left, and you will see many options for analyses, including fitting polynomials. Take a look at the Correlation and Regression section of the One Page Guides for instructions and videos. 

For Analyze > Fit Model, you do not have the model set up correctly. You will want to specify your outcome variable as Y, but you will not use the By section, that's for when you want to fit the same model across subsets of your data defined by the By variable. Instead, your model effects go in the Model Effects section. If you add your X variable there alone, that's a linear model. If you cross your x variable with itself, that's the squared term for a 2nd order polynomial model. You can build a great variety of models here, and if you are interested in seeing some of that I suggest watching the Mastering JMP Webinar for Specifying and Fitting Models. That webinar will use standard JMP, but you'll find that all of the options covered are in Student Edition -- Student Edition is incredibly powerful and, frankly, not that limited because we agree, as you suggested, that students need to have access to powerful tools in order to learn. 

I hope this helps get you started!

Julian

[jjproctor]

Julian:

Thank you very much for your help.  You were correct, when I imported the data there were words or letters in some of the columns, and so JMP set not just the modeling type of Nominal, but the Data type of the column to Character. I imported just the numerical values and everything worked as expected. 

We can close this request.

Thanks again,

John

[julian]

Hi @jjproctor,

I'm so happy to hear that everything worked as expected! We're all here to help if you hit any roadblocks, so never hesitate to reach out. 

Julian