src/Gargantext/Viz/Phylo/Aggregates/Fis.hs

   1 {-|
   2 Module      : Gargantext.Viz.Phylo.Tools
   3 Description : Phylomemy Tools to build/manage it
   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 {-# LANGUAGE NoImplicitPrelude #-}
  14 {-# LANGUAGE FlexibleContexts  #-}
  15 {-# LANGUAGE OverloadedStrings #-}
  16
  17 module Gargantext.Viz.Phylo.Aggregates.Fis
  18   where
  19
  20 import Data.List        (head)
  21 import Data.Map         (Map)
  22 import Data.Text        (words)
  23 import Data.Tuple       (fst, snd)
  24 import Gargantext.Prelude                       hiding (head)
  25 import Gargantext.Text.Metrics.FrequentItemSet  (fisWithSizePolyMap, Size(..))
  26 import Gargantext.Viz.Phylo
  27 import Gargantext.Viz.Phylo.Tools
  28 import qualified Data.Map    as Map
  29
  30
  31 -- | To Filter Fis by support
  32 filterFisBySupport :: Bool -> Int -> Map (Date, Date) [Fis] -> Map (Date, Date) [Fis]
  33 filterFisBySupport empty min' m = case empty of
  34   True  -> Map.map (\l -> filterMinorFis min' l) m
  35   False -> Map.map (\l -> keepFilled (filterMinorFis) min' l) m
  36
  37
  38 -- | To filter Fis with small Support, to preserve nonempty periods please use : filterFisBySupport False
  39 filterMinorFis :: Int -> [Fis] -> [Fis]
  40 filterMinorFis min' l = filter (\fis -> snd fis > min') l
  41
  42
  43 -- | To filter nested Fis
  44 filterFisByNested :: Map (Date, Date) [Fis] -> Map (Date, Date) [Fis]
  45 filterFisByNested = map (\l -> let cliqueMax = filterNestedSets (head $ map fst l) (map fst l) []
  46                                in  filter (\fis -> elem (fst fis) cliqueMax) l)
  47
  48
  49 -- | To transform a list of Documents into a Frequent Items Set
  50 docsToFis :: Map (Date, Date) [Document] -> Map (Date, Date) [Fis]
  51 docsToFis docs = map (\d -> Map.toList
  52                           $ fisWithSizePolyMap (Segment 1 20) 1 (map (words . text) d)) docs