I am creating a cross-lagged panel data model using the user-contributed package -xtdpdml-. My model is similar to the example specified in The Stata Journal (2018) 18, Number 2, pp. 293–326 Linear dynamic panel-data estimation using maximum likelihood and structural equation modeling.

Unlike the example in SJ, I have data at only 2 time points (T=2). This is my syntax:

Code:
xtdpdml outcome, pre(L.exposure) inv(cald) showcmd tfix
This results in the following error:

Code:
T value is too small given the lags specified
For example if xlag = 2 then T must equal at least 4
I have looked at the source code and I'm unsure of what "xlag" means in a practical sense. In my case, by using "set trace on", I find that T=2 (which I know from my data) and Tmin=3.

This leads to three questions:

(A) What does "xlag" mean?
(B) How is "Tmin" calculated? Why is this the minimum number of data points?
(C) Is it possible to create a cross-lagged panel model using -xtdpdml- with only T=2? Or do I need more time points?

The model runs fine when I use similar datasets with T=5 and T=7.