For data regression, how should these parameters be set?
Thread Content
This post was last edited by goat on 2010-4-8 at 15:16. I have been working on using Aspen’s data regression function to analyze experimental data, but there are many aspects I don’t understand and I’m not sure how to set them up; therefore, I’d like to discuss this with everyone. After selecting the regression method earlier on, one arrives at the “Parameters to be Regressed” screen, where there are options such as the TYPE option and the Name/Element option. For each type, there are often multiple Name options – how should one choose? Also, for some names, numbers such as 1, 1, 2, 2, 3 are required afterward. How can we determine how many numbers there are for a particular property method? Do anyone have any good textbooks on this topic? There is also the following Usage option: should you choose Regress or Fix?
1. Click on name/element and press F1 for helpful explanations. It explains in great detail the meaning of various binary parameters. 2. You can also specify whether the parameters you defined will be used in the regression, kept at the value entered in the Initial value field, or excluded from the regression. Choose regress, and your original values will be used for this regression ； Fix: determine the original value as the final value ； Exclude: the parameter values for regression cannot be the original values.
Everyone is welcome to share their own views and opinions! Communicate and discuss more
I’ve learned it, haha. I’m also quite confused about this issue
I’m confused too! ! Please give me some advice
Reply to 2# llbear: I have a question for fellow users. I have also been working on data fitting using UNIFAC, and there are several issues that have puzzled me during this process. 1. When using UNIFAC for fitting, do you need to enter volume and area parameters? I tried both approaches, but the results were inconsistent. 2 The results obtained after fitting differ significantly from those in the literature. What’s going on? The standard deviation of the regression results is quite large; how should I interpret this standard deviation? Thank you?