Updated on 2022-08-08 GMT+08:00

PageRank

Table 1 Parameter description

Parameter

Mandatory

Description

Type

Value Range

Default Value

alpha

No

Weight coefficient

(also called damping coefficient)

Double

A real number between 0 and 1 (excluding 0 and 1)

0.85

convergence

No

Convergence

Double

A real number between 0 and 1 (excluding 0 and 1)

0.00001

max_iterations

No

Maximum iterations

Integer

1 to 2000

1000

directed

No

Whether to consider the edge direction

Boolean

true or false

true

Iterations and convergence

The algorithm is terminated when either the maximum number of iterations is reached or the convergence precision is met.

  1. Generally, a smaller convergence precision and larger number of iterations lead to a better effect of the algorithm.
  2. To meet a certain convergence precision, you should set the number of iterations as large as possible.
  3. A larger number of iterations means a longer algorithm running time. To ensure that the algorithm runs at a certain number of iterations (that is, in a fixed duration), you should set the convergence precision as small as possible.
Table 2 response_data parameter description

Parameter

Type

Description

pagerank

List

PageRank value of each vertex. The format is as follows:

[{vertexId:rankValue},...],

where

vertexId is of the string type.

rankValue is of the double type.