mhahsler / dbscan

Density Based Clustering of Applications with Noise (DBSCAN) and Related Algorithms - R package

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

LOF return Nan values

isamsing opened this issue · comments

I have been using LOF for analysing more than 10,000 data point. However, the lof function in dbscan package return nan value.

Can you share a (small) dataset that produces NAN values with me?

Hello Micheal,
Thank for looking into the issue. I have attached the .csv file with this email.
Regards,Samneet Singh

Date: Thu, 31 Mar 2016 09:25:12 -0700
From: notifications@github.com
To: dbscan@noreply.github.com
CC: samneetdhillon@hotmail.com
Subject: Re: [mhahsler/dbscan] LOF return Nan values (#2)

Can you share a (small) dataset that produces NAN values with me?


You are receiving this because you authored the thread.
Reply to this email directly or view it on GitHub

I seem to not be able to see your attachment. Can you email it directly to me at michael at hahsler.net?

The problem was that the data has more than k duplicate points. The new version returns now 1 instead of NaN. However, the LOF paper suggests to first remove duplicates before running LOF.