I have GPS coordinates (latitude and longitude) for players (each with their own column in the matrix) on a team over a 60 second period (i.e t1 = [0:60] and I am trying to find the correlation coefficient for displacement for certain pairs of players (i.e. player 1 with player 2, player 5 with player 9, etc.) for the entire 60 second period. However, at certain time points, some players have an NaN value due to a weak signal. How can I calculate an r-value for each pair of players while excluding NaN values so I don't get an NaN value my r-value?