src/Gargantext/Text/Metrics/TFICF.hs

   1 {-|
   2 Module      : Gargantext.Text.Metrics.TFICF
   3 Description : TFICF Ngrams tools
   4 Copyright   : (c) CNRS, 2017
   5 License     : AGPL + CECILL v3
   6 Maintainer  : team@gargantext.org
   7 Stability   : experimental
   8 Portability : POSIX
   9
  10 Definition of TFICF : Term Frequency - Inverse of Context Frequency
  11
  12 TFICF is a generalization of [TFIDF](https://en.wikipedia.org/wiki/Tf%E2%80%93idf).
  13
  14 -}
  15
  16
  17 module Gargantext.Text.Metrics.TFICF ( TFICF
  18                                      , TficfContext(..)
  19                                      , Total(..)
  20                                      , Count(..)
  21                                      , tficf
  22                                      , sortTficf
  23                                      )
  24   where
  25
  26 import qualified Data.List as List
  27 import Data.Map.Strict (Map, toList)
  28 import qualified Data.Ord as DO (Down(..))
  29 import Data.Set (Set)
  30 import Data.Text
  31
  32 import Gargantext.Prelude
  33 import Gargantext.Core.Types (Ordering(..))
  34
  35 path :: Text
  36 path = "[G.T.Metrics.TFICF]"
  37
  38 type TFICF = Double
  39
  40 data TficfContext n m = TficfInfra n m
  41                       | TficfSupra n m
  42   deriving (Show)
  43
  44 data Total = Total {unTotal :: !Double}
  45 data Count = Count {unCount :: !Double}
  46
  47 tficf :: TficfContext Count Total
  48       -> TficfContext Count Total
  49       -> TFICF
  50 tficf (TficfInfra (Count ic) (Total it) )
  51       (TficfSupra (Count sc) (Total st) )
  52             | it >= ic && st >= sc {-&& it <= st-} = (ic/it) / log (sc/st)
  53             | otherwise            = panic $ "[ERR]" <> path <>" Frequency impossible"
  54 tficf _ _ = panic $ "[ERR]" <> path <> "Undefined for these contexts"
  55
  56
  57 sortTficf :: Ordering
  58           -> (Map Text (Double, Set Text))
  59           -> [   (Text,(Double, Set Text))]
  60 sortTficf Down = List.sortOn (DO.Down . fst . snd) . toList
  61 sortTficf Up   = List.sortOn (fst . snd) . toList
  62