1 unstable release

0.0.0-a5 Jun 20, 2024
0.0.0-a1 Jun 15, 2024
0.0.0-a0 May 17, 2024

#1739 in Data structures

50 downloads per month
Used in pyo3-graphster

BSD-3-Clause

150KB
2.5K SLoC

Graphster logo

Graphs without the headache: simple, powerful, effective

Graphster is a DataGraph library designed for rapid data manipulation, analysis, and processing


lib.rs:

Graphster

Graphster is a versatile graph data structure library designed for rapid data manipulation, analysis, and processing. It provides a rich set of features to create, manage, and manipulate graphs with various types of nodes and edges, each capable of holding multiple attributes.

Features

  • Nodes and Edges: Add, remove, and manipulate nodes and edges with ease.
  • Attributes: Each node and edge can have associated attributes with flexible types, including integers, floats, strings, and more.
  • Error Handling: Comprehensive error handling for common graph operations.
  • Polars Integration: Optional support for Polars data frames to import graph data.

Examples

Creating a Graph

use graphster::prelude::*;
use std::collections::HashMap;

let mut graph = DataGraph::new();

// Add nodes with attributes
graph.add_node(0, Attributes::from([("foo", "bar")])).unwrap();
graph.add_node(1, HashMap::from([("bar", "foo")])).unwrap();

// Add an edge between nodes
graph.add_edge(0, 1, Attributes::from([("weight", 10)])).unwrap();

// Get node and edge counts
assert_eq!(graph.node_count(), 2);
assert_eq!(graph.edge_count(), 1);

Using Attributes

use graphster::prelude::*;

let mut attrs = Attributes::new();
attrs.insert("key1", "value1");
attrs.insert("key2", 42);

assert_eq!(attrs.get("key1"), Some(&AttributeValue::String("value1".to_string())));
assert_eq!(attrs.get("key2"), Some(&AttributeValue::Int32(42)));

Handling Errors

use graphster::prelude::*;

let mut graph = DataGraph::new();
match graph.add_edge(0, 1, Attributes::new()) {
    Ok(_) => println!("Edge added successfully"),
    Err(e) => println!("Error adding edge: {}", e),
}

Using Polars DataFrames

use graphster::prelude::*;
use polars::prelude::*;

let s0 = Series::new("index", &[0, 1, 2]);
let s1 = Series::new("values", &[Some(1), None, Some(3)]);
let df = DataFrame::new(vec![s0, s1]).unwrap();

let nodes_df = NodesDataFrame::new(&df, "index");
let graph = DataGraph::from_nodes_dataframes(nodes_df).unwrap();

Examples of Advanced Operations

Neighbors

Retrieve all neighboring nodes that can be reached by outgoing edges from a specified node.

use graphster::prelude::*;

let mut graph = DataGraph::new();
graph.add_node(0, Attributes::new()).unwrap();
graph.add_node(1, Attributes::new()).unwrap();
graph.add_edge(0, 1, Attributes::new()).unwrap();

let neighbors: Vec<_> = graph.neighbors(0).unwrap().collect();
assert_eq!(neighbors, vec![&1.into()]);

Undirected Neighbors

Retrieve all neighboring nodes regardless of the direction of the edges.

use graphster::prelude::*;

let mut graph = DataGraph::new();
graph.add_node(0, Attributes::new()).unwrap();
graph.add_node(1, Attributes::new()).unwrap();
graph.add_edge(0, 1, Attributes::new()).unwrap();
graph.add_edge(1, 0, Attributes::new()).unwrap();

let neighbors: Vec<_> = graph.neighbors_undirected(0).unwrap().collect();
assert_eq!(neighbors, vec![&1.into()]);

Edges Connecting Nodes

Retrieve all edges connecting a specified source node to a target node.

use graphster::prelude::*;

let mut graph = DataGraph::new();
graph.add_node(0, Attributes::new()).unwrap();
graph.add_node(1, Attributes::new()).unwrap();
let edge_index = graph.add_edge(0, 1, Attributes::new()).unwrap();

let edges: Vec<_> = graph.edges_connecting(0, 1).unwrap().collect();
assert_eq!(edges, vec![&edge_index]);

Undirected Edges Connecting Nodes

Retrieve all edges connecting two nodes, regardless of the direction of the edges.

use graphster::prelude::*;

let mut graph = DataGraph::new();
graph.add_node(0, Attributes::new()).unwrap();
graph.add_node(1, Attributes::new()).unwrap();
let edge_index_1 = graph.add_edge(0, 1, Attributes::new()).unwrap();
let edge_index_2 = graph.add_edge(1, 0, Attributes::new()).unwrap();

let edges: Vec<_> = graph.edges_connecting_undirected(0, 1).unwrap().collect();
assert_eq!(edges, vec![&edge_index_1, &edge_index_2]);

Optional Features

  • Polars: Enable support for Polars data frames by specifying the polars feature.

Dependencies

~2–11MB
~112K SLoC