AVE is pretty straightforward. Here's a link to an excel toolkit that will help you calculate the values, based on lavaan output: https://www.scribd.com/document/384347064/AVE-and-Composite-Reliability-Calculator
David Morse Thank you for your informative response. I was interested to calculate AVE in R, so that the results can be easily replaceable by just running codes. But, it seems like most of researchers do it manually or with excel sheet. Now, I will gladly follow the links you shared.
I was going to suggest that you could code the formula in R without too much trouble. Seriously, I'm a newbie with R too. But then why start from scratch when you can use existing code?
This online exchange on RHelp suggests that AVE is can be reported in Lavaan. I think it refers to an earlier version of Lavaan, but should still apply.
Hume F. Winzar Thanks for your suggestions. I actually checked the reliability function of lavaan to calculate the AVE but the results were slightly different and in case where one latent has two indicators; the results have much difference. Hence, I just coded the formula as per my model requirement for both AVE and C.R. So that, I can re-run my model in future.
You can use the reliability function in semTools package to calculate the average variance extracted (AVE) and composite reliability (CR): semTools::reliability (your_model)
Here is a function (condisc) I have programmed in R that examines the convergent and discriminant validity of a measurement model estimated using the cfa()-function of lavaan: https://github.com/mmoglu/condisc