Normalizing Data to Initial Value

[#1]

I have a data set that I am plotting against time, and I was wondering if there was a feature in JMP to normalize the data to the initial value when graphing using the Bivariate function. I can do this by manually by creating a new column and calculating the normalized value, but I was hoping there was a way to normalize the data from the Bivariate menu instead. I've attached a screenshot of a sample data set to show what I am trying to do. Every value is substracted from the initial value. 

[#2 · Accepted solution]

The only way that I can think of to handle your request in the Bivariate Platform, is to create a Transform variable in the Bivariate Dialog Box. This would basically be the same as creating a new column in the data table, except that it would be a virtual column that exists only for the running of the Bivariate.

[#3]

Thanks for your clarification, I was hoping there would be a built-in option like 'Standardize'. However, I suppose I can just write a script to create the normalized columns and then perform a bivariate analysis. Thank you for the help!

[#4]

Hi @DanAlexander ,

This would be a good request to put on the JMP Wish List. In fact, this type of normalization has a name: Change from Baseline. One picks what baseline is (which row or measurement) and then performs the subtraction as you have done. This is common among clinical trials along with % change from baseline. Usually it is done per subject, but applies to many fields, could even be good for part number or other item where you track changes across time. MANOVA is a good test for statistical changes across time as well.

Another way people look a this is using time as a baseline covariate where time 0 is the reference and then model the response relative to some other fixed factor (also found in clinical trials and genomic studies).

Cheers,

[#5]

To add to @Chris_Kirchberg 's comment on this common practice in clinical trials, this indexing of a variable is also very common in econometrics. Many macro economic statistics use a specific value at a specific point in time as = 100. Then each subsequent value in the time series is an index off of that baseline value. Very very common...measures like Consumer Price Index, Industrial Production Index, etc. are all of this type. So to reinforce @Chris_Kirchberg 's idea, putting this in the JMP User Community Wish List is a great way to suggest to JMP product management and development your desire. Here's the landing page link for the JMP User Community Wish List: JMP User Community Wish List The other beauty of the Wish List is other JMP users can vote on the idea...and I know JMP product management and development take the votes to heart!

[#6]

I published an Addin : Normalization GUI .
The GUI facilitates the process of normalizing data to reference values, e.g. 0h data.