Fitting linear log regression with fitlm function

I would like to fit y = log(x) + k using fitlm (linear regression). How can I specify such a model? This is linear regression so would like to use fitlm function and not the fitnlm. Thanks!

Answer 1

[Star Strider]

Try this:

mdl = fitlm(log(x), y);

It assumes a linear model (slope and intercept), so this is all that is necessary.

NOTE — This will work providing that ‘x’ is real and positive. (I mention that for clarity.)

[Spirit]

Thanks StarStrider. I could do this way but this prevents me from using standard fitlm postprocessing functions say plot(mdl) for example. It plots a linear function as the actual regression is happening on log(x) vs. y.

I'd like to get a regression fit to y ~ log(x) + y using fitlm. I'd this should be possible as this still falls under the realm of linear regression.

Best, Shriram

[Star Strider]

I’m lost.

Good luck!