# Loglinear regression vs logarithmic regression

does anyone know if these two are the same thing? or if theyre different, could you explain what the difference is? thanks!

I believe they are the same thing.

i think these two are different. log-linear is when you turn a log equation into a linear equation. i don’t think you can use ordinary least squares on a logarithmic equation.

what about logistic regression? i’d like to see a 3-way comparison!

I think log linear means one of your axis (Y axis preferably) is log of linear variable and other axis is the linear variable, can be used if the relationship is logarithmic. Logarithmic means, you’re fitting logarithmic relationship model, so results will be same as log linear, but it’ll look different on chart, former would look linear and latter would look logarithmic. That’s my guess, I’m not sure at all :)!

Regressions are linear by definition. So, it doesn’t make sens to have a non-linear “logarithmic regression”. I’m pretty sure this just means that it’s a log-linear regression.