GitHub user njayaram2 opened a pull request:
https://github.com/apache/incubator-madlib/pull/137
Bugfix: Pagerank uses hard coded default threshold
JIRA: MADLIB-1100
The default threshold valud is set to 1e-5 instead of a value
dependent on the number of vertices (1/num_vertices*1000). The
older docs said it would be (1/number_vertices*100), but latest
experiments showed even that would converge in about 5 iterations.
So using 1000 instead of 100 now.
This commit also distributes intermediate and temp edge table by
both grouping cols and the dest vertex, hoping it would perform
better with MPP databases such as Greenplum and HAWQ.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/njayaram2/incubator-madlib
bugfix/pagerank_grouping
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/incubator-madlib/pull/137.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #137
----
commit 823e216a9dae22a60e0d25c0b1d9f3208ab9fcef
Author: Nandish Jayaram <[email protected]>
Date: 2017-05-19T18:00:03Z
Bugfix: Pagerank uses hard coded default threshold
JIRA: MADLIB-1100
The default threshold valud is set to 1e-5 instead of a value
dependent on the number of vertices (1/num_vertices*1000). The
older docs said it would be (1/number_vertices*100), but latest
experiments showed even that would converge in about 5 iterations.
So using 1000 instead of 100 now.
This commit also distributes intermediate and temp edge table by
both grouping cols and the dest vertex, hoping it would perform
better with MPP databases such as Greenplum and HAWQ.
----
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
