SimplePageRankVertex is just a simple example. Does this patch provide anything in addition to PageRankVertex?
The formula there is: double rank = (1d - teleportationProbability) * (rankFromNeighbors + danglingContribution) + teleportationProbability / getTotalNumVertices(); Cheers, -- Gianmarco On 23 June 2014 21:09, Pavan Kumar (JIRA) <j...@apache.org> wrote: > > [ > https://issues.apache.org/jira/browse/GIRAPH-877?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14041140#comment-14041140 > ] > > Pavan Kumar commented on GIRAPH-877: > ------------------------------------ > > the provided patch uses the correct equation (accounting for sink) & is > production tested > double value = d * sum + (d * sinkSum / n) + (1.0 - d) * allSum / n; > > in SimplePageRankComputation the formula used is > double value = d * sum + (1.0 - d) / n; > so it does not account for weights, etc. > > > pagerank application on giraph > > ------------------------------ > > > > Key: GIRAPH-877 > > URL: https://issues.apache.org/jira/browse/GIRAPH-877 > > Project: Giraph > > Issue Type: New Feature > > Environment: This patch has implementations for pagerank, > residual pagerank, weightedpagerank & residualweightedpagerank. > > Tests are also provided. mvn clean verify works. > > Reporter: Pavan Kumar > > Assignee: Pavan Kumar > > Attachments: pagerank.patch > > > > > > > > > -- > This message was sent by Atlassian JIRA > (v6.2#6252) >
