I am thinking maybe if I can do some equation on the variance of the coordinates.
EX: I sort the longitude column and add other column "RANK_X" holding sequence from 1 to the last row then I did the same for the Latitude adding "RANK_Y" column, and by summing those two rank columns "SUM_COORD" I have a number that indicates the variance in coordinates if I sorted this column the first 50 suppose to be near to each other based on the variance of the coordinates.
However, it's not the best approach and not giving the needed result in all the cases. see example below:
Case1:
Case2:
do you think I might get something from this approach maybe enhance the result by adding something to the equation?