7

I have a large amount of data that I want to cluster with Scikit's DBSCAN. I do it with the following line:

dbscanObject = DBSCAN(eps=20, min_samples=15).fit(featureVectors)

Unfortunately, this takes very long depending on how large the dataset is, and I see no way to monitor the progress of DBSCAN. How can I monitor the algorithm's progress and estimate the time left for the algorithm to finish?

Has QUIT--Anony-Mousse
  • 76,138
  • 12
  • 138
  • 194
PlsWork
  • 1,958
  • 1
  • 19
  • 31
  • If you can use keras wrappers look into [callbacks](https://keras.io/callbacks/) from fit function. Here is an [example repo.](https://github.com/stared/livelossplot) – Nihal Sangeeth May 20 '19 at 11:07
  • To add `callbacks` are currently not part of Scikit-learn but its part of the roadmap. Here's a [feature discussion on the same.](https://github.com/scikit-learn/scikit-learn/issues/10973) – Nihal Sangeeth May 20 '19 at 11:17
  • @NihalSangeeth It would be very helpful if you'd add an answer with a simple example with Keras wrapper. – Tharindu Sathischandra Nov 16 '20 at 12:02

0 Answers0