# Thread: Tips on how to regress? (non-linear)

1. ## Tips on how to regress? (non-linear)

Hey, everybody here always says that regressing is, like, the best thing the the whole world, but I find it extremely hard and tedious. Of course, I could just do it with my calculator, but I feel like that defeats the purpose. I assume when people say they regressed, that they were doing so non-linearly, as that's easy enough to to in your head. Perhaps everyone does logarithmic regressions, or exponential, or if they are lazy, quadratic? Could someone share some tips or mnemonics on the methods to regress some non-linear equations that they find satisfying?

All the best,
~Wyatt

2. Geez, I wish I could give more advice on this. The only time I learned to regress was back in my university days, and they just showed us how to do it using the computer. I've never experienced real-life regression

I don't know which of those regressions would create the best fit for you, but I think you're a bit of an outlier in terms of your needs. Hopefully that doesn't throw your correlation too much! It does sound like a significant question, and I wouldn't want to give you an error (type 1 or 2). I think someone else should take this one, because while I have a lot of deviations, this isn't one of the standard ones for me.

3. :p I mostly posted this as a joke, but thanks for your sincere reply!

4. Originally Posted by TeddyHugs
:p I mostly posted this as a joke, but thanks for your sincere reply!
Haha, my reply was a joke as well. I might be rustier on my stats terminology than I remember.

5. Oh I see. Clever girl...

6. Rather than a regression in calculation terms it's more like an exponential growth. Almost like a virus that sweeps over you... the little emerges and dominates.

7. This genuinely made me laugh out loud!