I am currently working with a difference-in-difference model specified as follows:


log(Viewsit)= B0 + σ0 (Cholerait) + B1 (Iraqit) + σ1 (Cholerait * Iraqtit) + B2 (Tabletit) + B3 (Computerit) + εit


My data includes daily views on a youtube animation, whether those views came from a tablet, cell phone or laptop, where those views originated from (with control group as Yemen and treatment group as Iraq). In 2015 there was a cholera epidemic in Iraq, such that the number of views shot up for then but quickly returned to the mean after the event. I have read that it is possible to do RD with time as a running variable (http://deep.ucdavis.edu/uploads/5/6/...deep_wp019.pdf). Would anyone know how to go about writing the Stata code for RD in this scenario? Thank you.