]> Git — Sourcephile - gargantext.git/blob - src/Gargantext/Core/Viz/Graph/Tools/IGraph.hs
[FEAT] connnecting new clustering algorithm
[gargantext.git] / src / Gargantext / Core / Viz / Graph / Tools / IGraph.hs
1 {-|
2 Module : Gargantext.Core.Viz.Graph.Tools.IGraph
3 Description : Tools to build Graph
4 Copyright : (c) CNRS, 2017-Present
5 License : AGPL + CECILL v3
6 Maintainer : team@gargantext.org
7 Stability : experimental
8 Portability : POSIX
9
10 Reference:
11 * Gábor Csárdi, Tamás Nepusz: The igraph software package for complex network research. InterJournal Complex Systems, 1695, 2006.
12
13 -}
14
15 module Gargantext.Core.Viz.Graph.Tools.IGraph
16 where
17
18 import Data.Serialize
19 import Data.Singletons (SingI)
20 import IGraph hiding (mkGraph, neighbors, edges, nodes, Node, Graph)
21 import Protolude
22 import Gargantext.Core.Viz.Graph.Index
23 import Graph.Types
24 import qualified Data.List as List
25 import qualified IGraph as IG
26 import qualified IGraph.Algorithms.Clique as IG
27 import qualified IGraph.Algorithms.Community as IG
28 import qualified IGraph.Algorithms.Structure as IG
29 import qualified IGraph.Random as IG
30 import qualified Data.Map as Map
31
32 ------------------------------------------------------------------
33 -- | Main Types
34 type Graph_Undirected = IG.Graph 'U () ()
35 type Graph_Directed = IG.Graph 'D () ()
36
37 type Node = IG.Node
38 type Graph = IG.Graph
39
40 ------------------------------------------------------------------
41 -- | Main Graph management Functions
42 neighbors :: IG.Graph d v e -> IG.Node -> [IG.Node]
43 neighbors = IG.neighbors
44
45 edges :: IG.Graph d v e -> [Edge]
46 edges = IG.edges
47
48 nodes :: IG.Graph d v e -> [IG.Node]
49 nodes = IG.nodes
50
51 ------------------------------------------------------------------
52 -- | Partitions
53 maximalCliques :: IG.Graph d v e -> [[Int]]
54 maximalCliques g = IG.maximalCliques g (min',max')
55 where
56 min' = 0
57 max' = 0
58
59 ------------------------------------------------------------------
60 type Seed = Int
61
62 spinglass :: Seed -> Map (Int, Int) Double -> IO [ClusterNode]
63 spinglass s g = toClusterNode
64 <$> map catMaybes
65 <$> map (map (\n -> Map.lookup n fromI))
66 <$> partitions_spinglass' s g'''
67 where
68 g' = toIndex toI g
69 g'' = mkGraphUfromEdges (Map.keys g')
70 g''' = case IG.isConnected g'' of
71 True -> g''
72 False -> case head (IG.decompose g'') of
73 Nothing -> panic "[G.C.V.G.T.Igraph: not connected graph]"
74 Just g'''' -> g''''
75
76 (toI, fromI) = createIndices g
77
78 -- | Tools to analyze graphs
79 partitions_spinglass' :: (Serialize v, Serialize e)
80 => Seed -> IG.Graph 'U v e -> IO [[Int]]
81 partitions_spinglass' s g = do
82 gen <- IG.withSeed s pure
83 IG.findCommunity g Nothing Nothing IG.spinglass gen
84
85
86 toClusterNode :: [[Int]] -> [ClusterNode]
87 toClusterNode ns = List.concat
88 $ map (\(cId, ns') -> map (\n -> ClusterNode n cId) ns')
89 $ List.zip [1..] ns
90
91 ------------------------------------------------------------------
92 mkGraph :: (SingI d, Ord v,
93 Serialize v, Serialize e) =>
94 [v] -> [LEdge e] -> IG.Graph d v e
95 mkGraph = IG.mkGraph
96
97 ------------------------------------------------------------------
98 mkGraphUfromEdges :: [(Int, Int)] -> Graph_Undirected
99 mkGraphUfromEdges es = mkGraph (List.replicate n ()) $ zip es $ repeat ()
100 where
101 (a,b) = List.unzip es
102 n = List.length (List.nub $ a <> b)
103
104 {-
105 mkGraphDfromEdges :: [(Int, Int)] -> Graph_Directed
106 mkGraphDfromEdges = undefined
107 -}