Need help with multivariate non-linear regression on a sizable dataset

Jan 2018
Zachary, LA
My dataset is attached (it's just a comma-delimited .txt file - change the extension to ".csv" to open in Excel). In the dataset, I have the following values:
  • dfn
  • h
  • p
  • s
  • ul

I need to form an equation that best fits my data and solves for "ul" (ul=?). This will not be a linear equation. I have ran linear regression on the dataset in Excel but the margin of error(?) was too large. I'm no Math wiz, I took several Calculus classes in college, but that was 15 years ago and I rarely use it.

I understand that in order to do multivariate non-linear regression on my data, I need to come up with a model equation. This is where my head explodes, I have no idea what this model would look like. Can someone maybe tell me what the model equation should be or at least point me in the right direction to determine it? Again, I don't speak high level Math, so try to put it in English for me (Wink)

View attachment


MHF Helper
Nov 2013
You must have taken this data for a reason. You must have some vague idea of the overall system model.

What generated this data? (I haven't looked at it, I don't trust unknown zip files)
Jan 2018
Zachary, LA
I replied to this but it's not showing up... ???