Hello Slim,
The short answer is yes and no. You may technically do this by using
vce(bootstrap, cluster(var)). However, this affects how the bootstrap is executed and you should think what this means in your case. Note that -npregress- is built for cross-sectional data so using panel data may be appropriate in some but not all cases.
P.S: Sorry for the late reply. Note you can always contact technical
support and will receive a more prompt answer. I happened to come to the
post by chance looking for wording I had used in the past.