SlopeSo I have to do an analysis of data where I have look at the change of a control over time by reagent masterlot and kit serial number. So I was wanting to look at the slope of the control by the reagent masterlot and kit serial. The table has a column for the result as well as columns that indicate the reagetmasterlot and the kit serial. So I want to be able to calculate the slope of the control result by reagentmasterlot and kit serial. Some controls will have a few data points and others a few more. The y would be the result value and the x the timedate stamp of that result. Any ideas on how to tackle this? I would put this analysis as a post script query as I am getting the data via a custom SQL from a database.
ScriptingRe: SlopeIf I have interpreted your question correctly, I believe you can easily get what you want by using
Analyze==>Screening==>Response Screening
You would specify your X and Y variables and then grouping variables as reagent masterlot and kit serial. Results of the Platform is a data table with statistics for each level, which includes the Slope.
Re: SlopeHi,
Not sure if I got it correctly, but you can make a "Fit Y by X" and put under "By" the reagent and kit to make the unique slopes (Untitled 10.jmp). Fit a line to all combination (shift and then the red triangle). Right click on parameter estimatesand select "make combined data table" witch gives you intercept and slope per reagent and kit (Untitled 30.jmp).
JSL for the last step: Make combined data table from two fit models
Re: SlopeThanks for the quick responses!
