src/Gargantext/Core/Text/Metrics/CharByChar.hs

   1 {-|
   2 Module      : Gargantext.Core.Text.Metrics.CharByChar
   3 Description : All parsers of Gargantext in one file.
   4 Copyright   : (c) CNRS, 2017 - present
   5 License     : AGPL + CECILL v3
   6 Maintainer  : team@gargantext.org
   7 Stability   : experimental
   8 Portability : POSIX
   9
  10 Mainly reexport functions in @Data.Text.Metrics@
  11 -}
  12
  13
  14
  15 module Gargantext.Core.Text.Metrics.CharByChar (levenshtein
  16                                       , levenshteinNorm
  17                                       , damerauLevenshtein
  18                                       , damerauLevenshteinNorm
  19                                       , overlap
  20                                       , jaccard
  21                                       , hamming
  22                                       ) where
  23
  24
  25 import Data.Text (Text)
  26 import GHC.Real (Ratio)
  27 import qualified Data.Text.Metrics as DTM
  28
  29 import Gargantext.Prelude
  30
  31 --noApax :: Ord a => Map a Occ -> Map a Occ
  32 --noApax m = M.filter (>1) m
  33
  34
  35 {- * Example de titre
  36 -}
  37
  38 -- | This module provide metrics to compare Text
  39 -- starting as an API rexporting main functions of the great lib
  40 -- text-metrics of Mark Karpov
  41
  42 -- | Levenshtein Similarity
  43 -- In information theory, Linguistics and computer science,
  44 -- the Levenshtein distance is a string metric for measuring
  45 -- the difference between two sequences.
  46 -- See: https://en.wikipedia.org/wiki/Levenshtein_distance
  47 --
  48 levenshtein :: Text -> Text -> Int
  49 levenshtein = DTM.levenshtein
  50
  51 -- | Return normalized Levenshtein distance between two 'Text' values.
  52 -- Result is a non-negative rational number (represented as @'Ratio'
  53 -- 'Data.Numeric.Natural'@), where 0 signifies no similarity between the
  54 -- strings, while 1 means exact match.
  55 --
  56 levenshteinNorm :: Text -> Text -> Ratio Int
  57 levenshteinNorm = DTM.levenshteinNorm
  58
  59 -- | Return Damerau-Levenshtein distance between two 'Text' values. The
  60 -- function works like 'levenshtein', but the collection of allowed
  61 -- operations also includes transposition of two /adjacent/ characters.
  62 -- See also:
  63 -- <https://en.wikipedia.org/wiki/Damerau%E2%80%93Levenshtein_distance>
  64 --
  65 damerauLevenshtein :: Text -> Text -> Int
  66 damerauLevenshtein = DTM.damerauLevenshtein
  67
  68 -- damerau-Levenshtein distance normalized
  69 --
  70 damerauLevenshteinNorm :: Text -> Text -> Ratio Int
  71 damerauLevenshteinNorm = DTM.damerauLevenshteinNorm
  72
  73 -- Treating inputs like sets
  74
  75 -- | Return overlap coefficient for two 'Text' values. Returned value
  76 -- is in the range from 0 (no similarity) to 1 (exact match). Return 1
  77 -- if both 'Text' values are empty.
  78 --
  79 -- See also: <https://en.wikipedia.org/wiki/Overlap_coefficient>.
  80 overlap :: Text -> Text -> Ratio Int
  81 overlap = DTM.overlap
  82
  83
  84 -- | Jaccard distance
  85 -- measures dissimilarity between sample sets
  86 jaccard :: Text -> Text -> Ratio Int
  87 jaccard = DTM.jaccard
  88
  89 -- | Hamming Similarity
  90 -- In information theory, the Hamming distance between two strings of
  91 -- equal length is the number of positions at which the corresponding
  92 -- symbols are different. In other words, it measures the minimum number of
  93 -- substitutions required to change one string into the other
  94 -- See:  https://en.wikipedia.org/wiki/Hamming_distance
  95
  96 hamming :: Text -> Text -> Maybe Int
  97 hamming = DTM.hamming
  98
  99