New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

More info regarding geocluster bias #3

Closed
kytwb opened this Issue Sep 7, 2017 · 2 comments

Comments

Projects
None yet
2 participants
@kytwb

kytwb commented Sep 7, 2017

I'm using the geocluster module to some map markers in our react native app but changing the bias parameter doesn't seem to change the geocluster function output (all my points are relatively close, all within 500m). In the README its set to 1.5.
Any hints on min/max value or anything else helpful? I'm trying to correlate the geocluster bias with what I have (latitude, longitude and deltas).

@yetzt

This comment has been minimized.

Show comment
Hide comment
@yetzt

yetzt Sep 7, 2017

Owner

the bias is not an absolute value, but a manipulator for the standard derivation of your data. there should be more clusters (because more points should be beyond the threshold for being in a new cluster, which is derived from the stddev) but it's dependent on the properties of the input.

(i wrote this quite some
time ago and don't have the theory present, but i shall revisit k-means, s-means and clustering at some
point.)

Owner

yetzt commented Sep 7, 2017

the bias is not an absolute value, but a manipulator for the standard derivation of your data. there should be more clusters (because more points should be beyond the threshold for being in a new cluster, which is derived from the stddev) but it's dependent on the properties of the input.

(i wrote this quite some
time ago and don't have the theory present, but i shall revisit k-means, s-means and clustering at some
point.)

@yetzt

This comment has been minimized.

Show comment
Hide comment
@yetzt

yetzt Sep 7, 2017

Owner

as far as i remember, all the *-means methods look for more or less equal size clusters, not density or other properties. so if this is useful to you depends on what you're trying to achieve.

Owner

yetzt commented Sep 7, 2017

as far as i remember, all the *-means methods look for more or less equal size clusters, not density or other properties. so if this is useful to you depends on what you're trying to achieve.

@yetzt yetzt closed this Sep 27, 2017

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment