Rsquared output from fitglm (logistic regression) - What does it mean?

Hello,

I'm using fitglm to fit a logistic regression model to some data:

mdl = fitglm(data,modelspec,'Distribution','binomial','CategoricalVars',[1]) % one categorical predictor

My question concerns the output of such a model.

What do the different fields in the mdl.Rsquared field mean and how are they computed? Are they different pseudo-Rsquares?

[MDias]

Any input would be very welcome.

[Kelly Hennigan]

I'd appreciate an answer to this question as well!

[Kelly Hennigan]

I just found a nice description of how each of the Rsquared values (output from fitglm) are calculated here:

https://www.mathworks.com/help/stats/generalizedlinearmodel.html

Looks like Rsquared.LLR is the McFadden pseudo R-squared for logistic regressions.