![]() XVals = J ( NRow (groupID ), 1, RandomNormal ()) GroupID = SortAscending ( Repeat (( 1 ::ng ) `, gs )) The code below took about 100 seconds on my laptop. If I need to do a script, I can do that as well.Ĭoincidentally, I was just about to point out that, if you know you need linear regression and only want the parameters, then JSL is much faster. Formulaįirst preference is to do with a formula. I know have something like this, and want a way to fill in the right 2 columns. I current have split my table by "Acid Level", splitting "pH", and keeping "Formula". I know that I can plot them in Fit Y by X, group the data, and then add linear fits but that 1) takes forever with 30k groups and 2) is a pain to pulls the slopes out of that report My data table is set up such as this, but with 30k Formula Combos and 13 acid levels for a total of 390k rows. ![]() For my range (and purpose) I can do this as a linear regression. ![]() I want to understand the slope of the pH as a function of acid concentration. I have 30k different formulations that have each been tested with 13 different levels of an acid. I want a column with slope for that row, and another with intercept for that row. I am trying to do a linear regression of 2 columns as grouped by a 3rd column and am stuck. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |