src/Gargantext/Pipeline.hs

   1 {-|
   2 Module      : Gargantext.Pipeline
   3 Description : Server API
   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 {-# OPTIONS_GHC -fno-warn-name-shadowing #-}
  13 {-# LANGUAGE NoImplicitPrelude           #-}
  14
  15 module Gargantext.Pipeline
  16   where
  17
  18 import Data.Text (unpack)
  19 import qualified Data.Text as DT
  20
  21 import Data.Text.IO (readFile)
  22
  23 ----------------------------------------------
  24 ----------------------------------------------
  25
  26 import Gargantext.Core
  27 import Gargantext.Core.Types
  28 import Gargantext.Prelude
  29
  30 import Gargantext.Viz.Graph.Index (map', createIndexes)
  31 import Gargantext.Viz.Graph.Distances.Matrice (distributional, int2double)
  32 import Gargantext.Text.Metrics.Occurrences
  33 import Gargantext.Text.Terms
  34 import Gargantext.Text.Context
  35
  36 import Data.Array.Accelerate as A
  37
  38 pipeline pth = do
  39   text     <- readFile pth
  40   let contexts = splitBy Sentences 4 text
  41   myterms <- mapM (terms Multi FR) contexts
  42   -- todo filter stop words
  43   let myCooc = removeApax $ cooc myterms
  44   --pure myCooc
  45   -- Cooc map -> Matrix
  46   --pure $ createIndexes myCooc
  47   pure $ map' int2double myCooc
  48   -- Matrix -> Graph
  49