Statistics: What is the difference between manova1() and anovan()?

I have basic knowledge in statistics and I don't know which function is the correct method for my data set.

I trained neural networks using 2 muscles groups separately (distal and proximal) to predict 2 outputs (angle and moment). In general, I want to know if using distal or proximal muscles gives me better predictions (aka statistically different).

I am using RMSE as a comparison metric.

Question - I don't know if I should have a single RMSE dependent variable (anovan) or 2: RMSE angle and RMSE moment (manova1) since it is the same metric but different units.

My factors are:

• Muscle groups: distal, proximal
• ?Dynamics: angle, moment
• walking condition: level, stair ascent, stair descent
• number of gait cycles in walking condition: 1 for level, 3 for SA, 3, for SD
• periods of gait cycle: load, mid stance, terminal stance, pre-swing, initial sw, mid sw, terminal sw

and I have 10 RMSE values per group. I have attached a figure that visually shows the data for a single muscle group. It shows an average of the 10 RMSE values.

I also want to know if at a specific period of a specific walking condition of a specific output is significantly different.

Question - Is that with multcompare()?

Question - can you do mulcompare() with stats from manova1()?

Question - Should I set up the data as a single column since I have 7 periods for level, 21 for stair ascent and 21 for stair descent or fill level with nan and do it as a matrix?

Thank you.