Water flows through a nanochannel and I have decided to count flowing oxygen atom in a nanotube. I used reduce/region command to sum the oxygen mass and then divide the result by the oxygen mass unit. Another way is to use the variable count(). However these two method do not lead into the same results.
Which one is more reliable?
It’s hard to say what is happening in your simulation if you don’t share the input files.
As a general rule of thumb, fork your simulation and try to reproduce this behaviour by increasingly simplifying the input, until a minimal working script is left. This helps a lot!
Neither. We have discussed this topic before, and what this is doing is measuring the occupancy and not the flow as it would be with the approaches described to your previously.
Why you get different results and whether you should or not is impossible to say without seeing your input.