Node degree from lat/lon

Hello, I'm trying to find the most popular locations from a list of 36000 lat/lon coordinates. I was hoping to use the indegree function and pull the top 10 results, but it doesnt seem to like to work with non-whole numbers such as lat/lon pairs. Is there a way of finding the most commonly occuring pairs from an array? It is currently set up as an x and y coordinate so would sort arrange them by x OR y, but not as a pair.

Thank you for your help.

Erick

[Steven Lord]

Do you have a digraph connecting the various locations specified by longitude and latitude values? Or do you just have a list of coordinates like:

rng default
xy = randi([1 10], 30, 2);

[Erick Koenig]

At the moment, it's a list of coordinates like [-74.1624 40.91464 ] in a 36000 X 2 array.

Answer 1

[Walter Roberson]

Is there a way of finding the most commonly occuring pairs from an array?

You do not have a graph, so you cannot use graph() object techniques.

One approach:

[count, ids] = groupcounts(YourMatrix);
NewMatrix = [ids{1}, ids{2}, count];

but... you should probably be using inexact comparisons, which groupcounts and findgroups will not do by default:

[groups, ~, groupnum] = uniquetol(YourMatrix, 'byrows', true);
count = histcounts(groupnum, 1:max(groupnum));
NewMatrix = [groups, count];

[Erick Koenig]

Thank you, I'll give that a try.

[Erick Koenig]

I'm getting..

Error using ismembertol

Too many input arguements

[Walter Roberson]

Sorry, I corrected the code.