]> Git — Sourcephile - gargantext.git/blob - src/Gargantext/Core/Viz/Graph/Tools.hs
[FIX] Map Text -> HashMap NgramsTerm
[gargantext.git] / src / Gargantext / Core / Viz / Graph / Tools.hs
1 {-|
2 Module : Gargantext.Core.Viz.Graph.Tools
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 -}
11
12
13 module Gargantext.Core.Viz.Graph.Tools
14 where
15
16 -- import Data.Graph.Clustering.Louvain (hLouvain, {-iLouvainMap-})
17 import Data.Graph.Clustering.Louvain.CplusPlus (cLouvain)
18 import Data.Graph.Clustering.Louvain.Utils (LouvainNode(..))
19 import Data.Map (Map)
20 import Data.HashMap.Strict (HashMap)
21 import Data.Text (Text)
22 import Debug.Trace (trace)
23 import GHC.Float (sin, cos)
24 import Gargantext.API.Ngrams.Types (NgramsTerm(..))
25 import Gargantext.Core.Methods.Distances (Distance(..), measure)
26 import Gargantext.Core.Methods.Graph.BAC.Proxemy (confluence)
27 import Gargantext.Core.Statistics
28 import Gargantext.Core.Viz.Graph
29 import Gargantext.Core.Viz.Graph.Bridgeness (bridgeness)
30 import Gargantext.Core.Viz.Graph.IGraph (mkGraphUfromEdges)
31 import Gargantext.Core.Viz.Graph.Index (createIndices, toIndex, map2mat, mat2map, Index)
32 import Gargantext.Prelude
33 import IGraph.Random -- (Gen(..))
34 import qualified Data.List as List
35 import qualified Data.Map as Map
36 import qualified Data.Set as Set
37 import qualified Data.Vector.Storable as Vec
38 import qualified IGraph as Igraph
39 import qualified IGraph.Algorithms.Layout as Layout
40 import qualified Data.HashMap.Strict as HashMap
41
42 type Threshold = Double
43
44
45 cooc2graph' :: Ord t => Distance
46 -> Double
47 -> Map (t, t) Int
48 -> Map (Index, Index) Double
49 cooc2graph' distance threshold myCooc = distanceMap
50 where
51 (ti, _) = createIndices myCooc
52 myCooc' = toIndex ti myCooc
53 matCooc = map2mat 0 (Map.size ti) $ Map.filter (> 1) myCooc'
54 distanceMat = measure distance matCooc
55 distanceMap = Map.filter (> threshold) $ mat2map distanceMat
56
57
58 cooc2graph :: Distance
59 -> Threshold
60 -> HashMap (NgramsTerm, NgramsTerm) Int
61 -> IO Graph
62 cooc2graph distance threshold myCooc = do
63 printDebug "cooc2graph" distance
64 let
65 -- TODO remove below
66 theMatrix = Map.fromList $ HashMap.toList myCooc
67 (ti, _) = createIndices theMatrix
68 myCooc' = toIndex ti theMatrix
69 matCooc = map2mat 0 (Map.size ti)
70 $ Map.filterWithKey (\(a,b) _ -> a /= b)
71 $ Map.filter (> 1) myCooc'
72 distanceMat = measure distance matCooc
73 distanceMap = Map.filter (> threshold) $ mat2map distanceMat
74
75 nodesApprox :: Int
76 nodesApprox = n'
77 where
78 (as, bs) = List.unzip $ Map.keys distanceMap
79 n' = Set.size $ Set.fromList $ as <> bs
80 ClustersParams rivers level = clustersParams nodesApprox
81
82
83 partitions <- if (Map.size distanceMap > 0)
84 -- then iLouvainMap 100 10 distanceMap
85 -- then hLouvain distanceMap
86 then cLouvain level distanceMap
87 else panic "Text.Flow: DistanceMap is empty"
88
89 let
90 -- bridgeness' = distanceMap
91 bridgeness' = trace ("Rivers: " <> show rivers)
92 $ bridgeness rivers partitions distanceMap
93 confluence' = confluence (Map.keys bridgeness') 3 True False
94
95 pure $ data2graph (Map.toList $ Map.mapKeys unNgramsTerm ti) myCooc' bridgeness' confluence' partitions
96
97
98
99 data ClustersParams = ClustersParams { bridgness :: Double
100 , louvain :: Text
101 } deriving (Show)
102
103 clustersParams :: Int -> ClustersParams
104 clustersParams x = ClustersParams (fromIntegral x) "0.00000001" -- y
105 {- where
106 y | x < 100 = "0.000001"
107 | x < 350 = "0.000001"
108 | x < 500 = "0.000001"
109 | x < 1000 = "0.000001"
110 | otherwise = "1"
111 -}
112
113 ----------------------------------------------------------
114 -- | From data to Graph
115 data2graph :: [(Text, Int)]
116 -> Map (Int, Int) Int
117 -> Map (Int, Int) Double
118 -> Map (Int, Int) Double
119 -> [LouvainNode]
120 -> Graph
121 data2graph labels coocs bridge conf partitions = Graph nodes edges Nothing
122 where
123
124 community_id_by_node_id = Map.fromList [ (n, c) | LouvainNode n c <- partitions ]
125
126 nodes = map (setCoord ForceAtlas labels bridge)
127 [ (n, Node { node_size = maybe 0 identity (Map.lookup (n,n) coocs)
128 , node_type = Terms -- or Unknown
129 , node_id = cs (show n)
130 , node_label = l
131 , node_x_coord = 0
132 , node_y_coord = 0
133 , node_attributes =
134 Attributes { clust_default = maybe 0 identity
135 (Map.lookup n community_id_by_node_id) } }
136 )
137 | (l, n) <- labels
138 , Set.member n $ Set.fromList
139 $ List.concat
140 $ map (\((s,t),d) -> if d > 0 && s /=t then [s,t] else [])
141 $ Map.toList bridge
142 ]
143
144 edges = [ Edge { edge_source = cs (show s)
145 , edge_target = cs (show t)
146 , edge_weight = d
147 , edge_confluence = maybe 0 identity $ Map.lookup (s,t) conf
148 -- , edge_confluence = maybe (panic "E: data2graph edges") identity $ Map.lookup (s,t) conf
149 , edge_id = cs (show i) }
150 | (i, ((s,t), d)) <- zip ([0..]::[Integer]) (Map.toList bridge), s /= t, d > 0
151 ]
152
153
154 ------------------------------------------------------------------------
155
156 data Layout = KamadaKawai | ACP | ForceAtlas
157
158
159 setCoord' :: (Int -> (Double, Double)) -> (Int, Node) -> Node
160 setCoord' f (i,n) = n { node_x_coord = x, node_y_coord = y }
161 where
162 (x,y) = f i
163
164
165 -- | ACP
166 setCoord :: Ord a => Layout -> [(a, Int)] -> Map (Int, Int) Double -> (Int, Node) -> Node
167 setCoord l labels m (n,node) = node { node_x_coord = x
168 , node_y_coord = y
169 }
170 where
171 (x,y) = getCoord l labels m n
172
173
174 getCoord :: Ord a
175 => Layout
176 -> [(a, Int)]
177 -> Map (Int, Int) Double
178 -> Int
179 -> (Double, Double)
180 getCoord KamadaKawai _ _m _n = undefined -- layout m n
181
182 getCoord ForceAtlas _ _ n = (sin d, cos d)
183 where
184 d = fromIntegral n
185
186 getCoord ACP labels m n = to2d $ maybe (panic "Graph.Tools no coordinate") identity
187 $ Map.lookup n
188 $ pcaReduceTo (Dimension 2)
189 $ mapArray labels m
190 where
191 to2d :: Vec.Vector Double -> (Double, Double)
192 to2d v = (x',y')
193 where
194 ds = take 2 $ Vec.toList v
195 x' = head' "to2d" ds
196 y' = last' "to2d" ds
197
198 mapArray :: Ord a => [(a, Int)] -> Map (Int, Int) Double -> Map Int (Vec.Vector Double)
199 mapArray items m' = Map.fromList [ toVec n' ns m' | n' <- ns ]
200 where
201 ns = map snd items
202
203 toVec :: Int -> [Int] -> Map (Int,Int) Double -> (Int, Vec.Vector Double)
204 toVec n' ns' m' = (n', Vec.fromList $ map (\n'' -> maybe 0 identity $ Map.lookup (n',n'') m') ns')
205 ------------------------------------------------------------------------
206
207 -- | KamadaKawai Layout
208 -- TODO TEST: check labels, nodeId and coordinates
209 layout :: Map (Int, Int) Double -> Int -> Gen -> (Double, Double)
210 layout m n gen = maybe (panic "") identity $ Map.lookup n $ coord
211 where
212 coord :: (Map Int (Double,Double))
213 coord = Map.fromList $ List.zip (Igraph.nodes g) $ (Layout.layout g p gen)
214 --p = Layout.defaultLGL
215 p = Layout.kamadaKawai
216 g = mkGraphUfromEdges $ map fst $ List.filter (\e -> snd e > 0) $ Map.toList m
217