1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121

// Copyright 2016 rust-tfidf Developers
//
// Licensed under the Apache License, Version 2.0, <LICENSE-APACHE or
// http://apache.org/licenses/LICENSE-2.0> or the MIT license <LICENSE-MIT or
// http://opensource.org/licenses/MIT>, at your option. This file may not be
// copied, modified, or distributed except according to those terms.

use std::borrow::Borrow;

use prelude::{NaiveDocument, NormalizationFactor, ProcessedDocument, Tf};

/// Binary weighting scheme for TF. If the document contains the term, returns 1,
/// otherwise returns 0.
#[derive(Copy, Clone)]
pub struct BinaryTf;

impl<T> Tf<T> for BinaryTf
where
  T: NaiveDocument,
{
  #[inline]
  fn tf<K>(term: K, doc: &T) -> f64
  where
    K: Borrow<T::Term>,
  {
    if doc.term_exists(term) {
      1f64
    } else {
      0f64
    }
  }
}

/// Raw frequency weighting scheme for TF. Returns the number of times a term occurs
/// in the document.
#[derive(Copy, Clone)]
pub struct RawFrequencyTf(f64);

impl<T> Tf<T> for RawFrequencyTf
where
  T: ProcessedDocument,
{
  #[inline]
  fn tf<K>(term: K, doc: &T) -> f64
  where
    K: Borrow<T::Term>,
  {
    doc.term_frequency(term) as f64
  }
}

/// Log normalized weighting scheme for TF. Computes `1 + log (f)` where `f` is the
/// frequency of the term in the document.
#[derive(Copy, Clone)]
pub struct LogNormalizationTf;

impl<T> Tf<T> for LogNormalizationTf
where
  T: ProcessedDocument,
{
  #[inline]
  fn tf<K>(term: K, doc: &T) -> f64
  where
    K: Borrow<T::Term>,
  {
    1f64 + (doc.term_frequency(term) as f64).ln()
  }
}

/// Double normalized weighting scheme for TF based on a factor, `K`.
///
/// # Example
///
/// To implement a custom Tf strategy, where the `K` factor is constant:
///
/// ```rust
/// use tfidf::{Tf, NormalizationFactor};
/// use tfidf::tf::{DoubleKNormalizationTf};
///
/// struct DoubleThirdNormalizationTf;
///
/// impl NormalizationFactor for DoubleThirdNormalizationTf {
///   fn factor() -> f64 { 0.3f64 }
/// }
///
/// impl DoubleKNormalizationTf for DoubleThirdNormalizationTf { }
/// ```
pub trait DoubleKNormalizationTf: NormalizationFactor {}

impl<T, S> Tf<T> for S
where
  S: DoubleKNormalizationTf,
  T: ProcessedDocument,
{
  #[inline]
  fn tf<K>(term: K, doc: &T) -> f64
  where
    K: Borrow<T::Term>,
  {
    let max = match doc.max() {
      Some(m) => doc.term_frequency(m) as f64,
      None => 1f64,
    };

    // K + ((1 - K) * (f / max f))
    S::factor() + ((1f64 - S::factor()) * ((doc.term_frequency(term) as f64) / max))
  }
}

/// Double normalized weighting scheme for TF based on a factor, `K = 0.5`.
#[derive(Copy, Clone)]
pub struct DoubleHalfNormalizationTf;

impl NormalizationFactor for DoubleHalfNormalizationTf {
  #[inline]
  fn factor() -> f64 {
    0.5f64
  }
}

impl DoubleKNormalizationTf for DoubleHalfNormalizationTf {}