-
-h --help: shows help message and exits [default: false] -
-v --version: prints version information and exits [default: false] -
-graph-path: Path of the graph file [required] -
-seed-set-path: Path of the seed set file [required] -
-algo: The algorithm to use:Auto,PR-IMM,SA-IMM,SA-RG-IMM,Greedy,MaxDegreeorPageRank[default:Auto] -
-k: Number of boosted nodes [required] -
-priority: Priority of the 4 states. Input from highest to lowest, tokens separated with spaces, commas or>. e.g.-priority Ca+ Cr- Cr Carefers to Ca+ > Cr- > Cr > Ca [required] -
-lambda: Weight parameter$\lambda$ of objective function [default: 0.5] -
-log-per-percentage: Frequency for progress logging [default: 5] -
-test-times: How many times to check the solution by forward simulation [default: 10000]
-k can be either a single positive integer (e.g. 10),
or a list of positive integers separated by spaces, commas or semicolons.
e.g. (10,20,30 or 10; 20; 30). The maximum value of k is used in the algorithm.
During simulation, however, all the k's are attempted
(since boosted nodes are sorted in descending order by their influence,
the result of boosted nodes with
PR-IMM algorithm supports two modes during sampling.
If the argument -n-samples is provided,
static sample set size is used where the number of PRR-sketch samples is a (series of) fixed value(s).
Otherwise, the sample set size
-n-samples: Fixed number of samples.
n-samples can be either a single positive integer, or a list of positive integers
separated by spaces, commas or semicolons. For two sample set sizes
-
-sample-limit: Maximum number of PRR-sketch samples [default: +inf] -
-epsilon:$\epsilon$ in PR-IMM algorithm [required] -
-ell:$\ell$ [default: 1]
SA-(RG-)IMM algorithm is performed as two parts, both of whose arguments shall be provided separately:
- Upper bound: assumes with a monotonic & submodular priority and performs PR-IMM algorithm;
- Lower bound (denoted as SA-(RG-)IMM-LB).
SA-(RG-)IMM-LB supports two modes during sampling:
If the argument -n-samples-sa is provided,
sample set size per center node (i.e. total number of samples = number of center nodes
Otherwise, sample set size per center
-sample-dist-limit-sa: Distance threshold from center nodes to any of the seeds. Nodes with higher distance are filtered out. [default: +inf]
-n-samples-sa: Fixed number of samples for each center node.
-n-samples-sa can be either a positive integer or a list of positive integers. See -n-samples in PR-IMM algorithm for details.
-
-sample-limit-sa: Limit of theta_sa in SA-(RG-)IMM algorithm [default: +inf] -
-epsilon-sa:$\epsilon_2$ in SA-(RG-)IMM algorithm [required] -
-ell:$\ell$ [default: 1]
-greedy-test-times: How many times to repeat per forward simulation during greedy algorithm. [default: 10000]