src/Gargantext/Database/Action/Metrics/TFICF.hs

   1 {-|
   2 Module      : Gargantext.Database.Metrics.TFICF
   3 Description : Ngrams by Node user and master
   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 {-# LANGUAGE QuasiQuotes       #-}
  13
  14 module Gargantext.Database.Action.Metrics.TFICF
  15   where
  16
  17 -- import Debug.Trace (trace)
  18 -- import Gargantext.Core (Lang(..))
  19 import Data.HashMap.Strict (HashMap)
  20 import qualified Data.HashMap.Strict as HM
  21 import Data.Maybe (fromMaybe)
  22 import Gargantext.Core.Text.Metrics.TFICF
  23 import Gargantext.Database.Action.Metrics.NgramsByNode (getNodesByNgramsUser, getOccByNgramsOnlyFast)
  24 import Gargantext.Database.Admin.Types.Node -- (ListId, CorpusId, NodeId)
  25 import Gargantext.Database.Prelude (Cmd)
  26 import Gargantext.Database.Query.Table.NodeNode (selectCountDocs)
  27 import Gargantext.Database.Schema.Ngrams (NgramsType(..))
  28 import Gargantext.API.Ngrams.Types
  29 import Gargantext.Prelude
  30 import qualified Data.Set as Set
  31
  32 getTficf :: UserCorpusId
  33          -> MasterCorpusId
  34          -> NgramsType
  35          -> Cmd err (HashMap NgramsTerm Double)
  36 getTficf cId mId nt = do
  37   mapTextDoubleLocal <- HM.filter (> 1)
  38      <$> HM.map (fromIntegral . Set.size)
  39      <$> getNodesByNgramsUser cId nt
  40
  41   mapTextDoubleGlobal <- HM.map fromIntegral
  42                      <$> getOccByNgramsOnlyFast mId nt (HM.keys mapTextDoubleLocal)
  43
  44   countLocal  <- selectCountDocs cId
  45   countGlobal <- selectCountDocs mId
  46
  47   pure $ HM.mapWithKey (\t n ->
  48       tficf (TficfInfra (Count n                                               )
  49                         (Total $ fromIntegral countLocal))
  50             (TficfSupra (Count $ fromMaybe 0 $ HM.lookup t mapTextDoubleGlobal)
  51                         (Total $ fromIntegral countGlobal))
  52     ) mapTextDoubleLocal