2 Module : Gargantext.Graph.Distances.Matrix
4 Copyright : (c) CNRS, 2017-Present
5 License : AGPL + CECILL v3
6 Maintainer : team@gargantext.org
7 Stability : experimental
10 Motivation and definition of the @Conditional@ distance.
12 Implementation use Accelerate library :
13 * Manuel M. T. Chakravarty, Gabriele Keller, Sean Lee, Trevor L. McDonell, and Vinod Grover.
14 [Accelerating Haskell Array Codes with Multicore GPUs][CKLM+11].
15 In _DAMP '11: Declarative Aspects of Multicore Programming_, ACM, 2011.
17 * Trevor L. McDonell, Manuel M. T. Chakravarty, Gabriele Keller, and Ben Lippmeier.
18 [Optimising Purely Functional GPU Programs][MCKL13].
19 In _ICFP '13: The 18th ACM SIGPLAN International Conference on Functional Programming_, ACM, 2013.
21 * Robert Clifton-Everest, Trevor L. McDonell, Manuel M. T. Chakravarty, and Gabriele Keller.
22 [Embedding Foreign Code][CMCK14].
23 In _PADL '14: The 16th International Symposium on Practical Aspects of Declarative Languages_, Springer-Verlag, LNCS, 2014.
25 * Trevor L. McDonell, Manuel M. T. Chakravarty, Vinod Grover, and Ryan R. Newton.
26 [Type-safe Runtime Code Generation: Accelerate to LLVM][MCGN15].
27 In _Haskell '15: The 8th ACM SIGPLAN Symposium on Haskell_, ACM, 2015.
31 {-# LANGUAGE NoImplicitPrelude #-}
32 {-# LANGUAGE FlexibleContexts #-}
33 {-# LANGUAGE TypeFamilies #-}
34 {-# LANGUAGE TypeOperators #-}
35 {-# LANGUAGE ScopedTypeVariables #-}
38 module Gargantext.Viz.Graph.Distances.Matrice
41 import Data.Array.Accelerate
42 import Data.Array.Accelerate.Interpreter (run)
43 import Data.Array.Accelerate.Smart
44 import Data.Array.Accelerate.Type
45 import Data.Array.Accelerate.Array.Sugar (fromArr, Array, Z)
47 import Data.Maybe (Maybe(Just))
48 import qualified Gargantext.Prelude as P
49 import qualified Data.Array.Accelerate.Array.Representation as Repr
51 import Gargantext.Text.Metrics.Count
54 -----------------------------------------------------------------------
56 distriTest = distributional $ myMat 100
57 -----------------------------------------------------------------------
59 vector :: Int -> (Array (Z :. Int) Int)
60 vector n = fromList (Z :. n) [0..n]
62 matrix :: Elt c => Int -> [c] -> Matrix c
63 matrix n l = fromList (Z :. n :. n) l
65 myMat :: Int -> Matrix Int
66 myMat n = matrix n [1..]
68 -- | Two ways to get the rank (as documentation)
69 rank :: (Matrix a) -> Int
70 rank m = arrayRank $ arrayShape m
72 -----------------------------------------------------------------------
73 -- | Dimension of a square Matrix
74 -- How to force use with SquareMatrix ?
77 dim :: (Matrix a) -> Dim
80 Z :. _ :. n = arrayShape m
81 -- == indexTail (arrayShape m)
83 -----------------------------------------------------------------------
84 proba :: Dim -> Acc (Matrix Double) -> Acc (Matrix Double)
85 proba r mat = zipWith (/) mat (mkSum r mat)
87 mkSum :: Dim -> Acc (Matrix Double) -> Acc (Matrix Double)
88 mkSum r mat = replicate (constant (Z :. (r :: Int) :. All)) $ sum mat
91 divByDiag :: Dim -> Acc (Matrix Double) -> Acc (Matrix Double)
92 divByDiag d mat = zipWith (/) mat (replicate (constant (Z :. (d :: Int) :. All)) $ diag mat)
94 diag :: Elt e => Acc (Matrix e) -> Acc (Vector e)
95 diag m = backpermute (indexTail (shape m)) (lift1 (\(Z :. x) -> (Z :. x :. (x :: Exp Int)))) m
96 -----------------------------------------------------------------------
98 miniMax :: Acc (Matrix Double) -> Acc (Matrix Double)
99 miniMax m = map (\x -> ifThenElse (x > miniMax') x 0) m
101 miniMax' = (the $ minimum $ maximum m)
103 -- | Conditional distance (basic version)
104 conditional :: Matrix Int -> Matrix Double
105 conditional m = run (miniMax $ proba (dim m) $ map fromIntegral $ use m)
108 -- | Conditional distance (advanced version)
109 conditional' :: Matrix Int -> (Matrix InclusionExclusion, Matrix SpecificityGenericity)
110 conditional' m = (run $ ie $ map fromIntegral $ use m, run $ sg $ map fromIntegral $ use m)
112 ie :: Acc (Matrix Double) -> Acc (Matrix Double)
113 ie mat = map (\x -> x / (2*n-1)) $ zipWith (+) (xs mat) (ys mat)
114 sg :: Acc (Matrix Double) -> Acc (Matrix Double)
115 sg mat = map (\x -> x / (2*n-1)) $ zipWith (-) (xs mat) (ys mat)
123 xs :: Acc (Matrix Double) -> Acc (Matrix Double)
124 xs mat = zipWith (-) (proba r mat) (mkSum r $ proba r mat)
125 ys :: Acc (Matrix Double) -> Acc (Matrix Double)
126 ys mat = zipWith (-) (proba r mat) (mkSum r $ transpose $ proba r mat)
128 -----------------------------------------------------------------------
130 -- | Distributional Distance
131 distributional :: Matrix Int -> Matrix Double
132 distributional m = run $ miniMax $ ri (map fromIntegral $ use m)
136 filter m = zipWith (\a b -> max a b) m (transpose m)
138 ri mat = zipWith (/) mat1 mat2
140 mat1 = mkSum n $ zipWith min (mi mat) (mi $ transpose mat)
143 mi m' = zipWith (\a b -> max (log $ a/b) 0) m'
144 $ zipWith (/) (crossProduct m') (total m')
146 total m'' = replicate (constant (Z :. n :. n)) $ fold (+) 0 $ fold (+) 0 m''
148 crossProduct m = zipWith (*) (cross m ) (cross (transpose m))
149 cross mat = zipWith (-) (mkSum n mat) (mat)
151 -----------------------------------------------------------------------
152 -----------------------------------------------------------------------
155 Metric Specificity and genericity: select terms
160 Probability to get i given j : P(i|j)=Nij/Nj
161 Gen(i) : 1/(N-1)*Sum(j!=i, P(i|j)) : Genericity of i
162 Spec(i) : 1/(N-1)*Sum(j!=i, P(j|i)) : Specificity of j
163 Inclusion (i) = Gen(i)+Spec(i)
164 Genericity score = Gen(i)- Spec(i)
168 * Science mapping with asymmetrical paradigmatic proximity Jean-Philippe Cointet (CREA, TSV), David Chavalarias (CREA) (Submitted on 15 Mar 2008), Networks and Heterogeneous Media 3, 2 (2008) 267 - 276, arXiv:0803.2315 [cs.OH]
172 type InclusionExclusion = Double
173 type SpecificityGenericity = Double
175 data SquareMatrix = SymetricMatrix | NonSymetricMatrix
176 type SymetricMatrix = Matrix
177 type NonSymetricMatrix = Matrix
180 incExcSpeGen :: Matrix Int -> (Vector InclusionExclusion, Vector SpecificityGenericity)
181 incExcSpeGen m = (run' inclusionExclusion m, run' specificityGenericity m)
183 run' fun mat = run $ fun $ map fromIntegral $ use mat
185 -- | Inclusion (i) = Gen(i)+Spec(i)
186 inclusionExclusion :: Acc (Matrix Double) -> Acc (Vector Double)
187 inclusionExclusion mat = zipWith (+) (pV mat) (pH mat)
189 -- | Genericity score = Gen(i)- Spec(i)
190 specificityGenericity :: Acc (Matrix Double) -> Acc (Vector Double)
191 specificityGenericity mat = zipWith (-) (pV mat) (pH mat)
193 -- | Gen(i) : 1/(N-1)*Sum(j!=i, P(i|j)) : Genericity of i
194 pV :: Acc (Matrix Double) -> Acc (Vector Double)
195 pV mat = map (\x -> (x-1)/(cardN-1)) $ sum $ p_ij mat
197 -- | Spec(i) : 1/(N-1)*Sum(j!=i, P(j|i)) : Specificity of j
198 pH :: Acc (Matrix Double) -> Acc (Vector Double)
199 pH mat = map (\x -> (x-1)/(cardN-1)) $ sum $ p_ji mat
202 cardN = constant (P.fromIntegral (dim m) :: Double)
205 -- | P(i|j) = Nij /N(jj) Probability to get i given j
206 p_ij :: (Elt e, P.Fractional (Exp e)) => Acc (SymetricMatrix e) -> Acc (Matrix e)
207 p_ij m = zipWith (/) m (n_jj m)
209 n_jj :: Elt e => Acc (SymetricMatrix e) -> Acc (Matrix e)
210 n_jj m = backpermute (shape m)
211 (lift1 ( \(Z :. (_ :: Exp Int) :. (j:: Exp Int))
216 -- | P(j|i) = Nij /N(ii) Probability to get i given j
218 p_ji :: (Elt e, P.Fractional (Exp e)) => Acc (Array DIM2 e) -> Acc (Array DIM2 e)
219 p_ji = transpose . p_ij
222 -- | Step to ckeck the result in visual/qualitative tests
223 incExcSpeGen_proba :: Matrix Int -> Matrix Double
224 incExcSpeGen_proba m = run' pro m
226 run' fun mat = run $ fun $ map fromIntegral $ use mat
231 -- | Hypothesis to test maybe later (or not)
232 -- TODO ask accelerate for instances to ease such writtings:
233 p_ :: (Elt e, P.Fractional (Exp e)) => Acc (Array DIM2 e) -> Acc (Array DIM2 e)
234 p_ m = zipWith (/) m (n_ m)
236 n_ :: Elt e => Acc (SymetricMatrix e) -> Acc (Matrix e)
237 n_ m = backpermute (shape m)
238 (lift1 ( \(Z :. (i :: Exp Int) :. (j:: Exp Int))
239 -> (ifThenElse (i < j) (lift (Z :. j :. j)) (lift (Z :. i :. i)) :: Exp DIM2)