So, I'm a R user and currently working on a spatial analysis on panel data. Based on my experience for these past days, I understand that
spml function on
splm package requires strict condition to actually run the analysis (e.g.: balanced panels and complete cases).
Recently, I tried to run
plm function (from pacakge
plm) instead with self-defined spatially-lagged dependent variable. Will this actually yield the same result? Is there important unaccounted spatial aspects if I use
plm instead of