Add customization of stemming function

README.md

@@ -36,6 +36,32 @@ A [Naive Bayes](https://en.wikipedia.org/wiki/Naive_Bayes_classifier) machine le
 
 ## Usage
 
+Install by adding `:simple_bayes` and optionally `:stemmer` to `deps` in your
+`mix.exs`:
+
+```elixir
+defp deps do
+  [ {:simple_bayes, "~> 0.10.0"},
+    {:stemmer, "~> 1.0"} # Optional, if you want to use stemming
+  ]
+end
+```
+
+Ensure `:simple_bayes` and optionally `:stemmer` are started before your
+application:
+
+```elixir
+def application do
+  [ applications: [
+      :logger,
+      :simple_bayes,
+      :stemmer # Optional, if you want to use stemming
+    ]
+  ]
+end
+```
+
+
 ```elixir
 bayes = SimpleBayes.init()
         |> SimpleBayes.train(:apple, "red sweet")
@@ -65,7 +91,7 @@ bayes |> SimpleBayes.classify("Maybe green maybe red but definitely round and sw
 # ]
 ```
 
-With and without word stemming:
+With and without word stemming (requires a stem function, we recommend [Stemmer](https://github.com/fredwu/stemmer)):
 
 ```elixir
 SimpleBayes.init()
@@ -77,7 +103,7 @@ SimpleBayes.init()
 #   apple: 0.05719389206673358
 # ]
 
-SimpleBayes.init(stem: true)
+SimpleBayes.init(stem: &Stemmer.stem/1) # Or any other stemming function
 |> SimpleBayes.train(:apple, "buying apple")
 |> SimpleBayes.train(:banana, "buy banana")
 |> SimpleBayes.classify("buy apple")
@@ -96,7 +122,7 @@ config :simple_bayes, model: :multinomial
 config :simple_bayes, storage: :memory
 config :simple_bayes, default_weight: 1
 config :simple_bayes, smoothing: 0
-config :simple_bayes, stem: false
+config :simple_bayes, stem: false # or a stemming function
 config :simple_bayes, top: nil
 config :simple_bayes, stop_words: ~w(
   a about above after again against all am an and any are aren't as at be
@@ -123,7 +149,7 @@ SimpleBayes.init(
   storage:        :memory,
   default_weight: 1,
   smoothing:      0,
-  stem:           false,
+  stem:           false, # or a stemming function
   top:            nil,
   stop_words:     []
 )

lib/simple_bayes/trainer/token_stemmer.ex

@@ -1,18 +1,15 @@
 defmodule SimpleBayes.Trainer.TokenStemmer do
   @doc """
-  Stems the word, or passes the word through.
+  Stems the word using `stemmer`. Passes the word through if it is false.
 
   ## Examples
 
-      iex> SimpleBayes.Trainer.TokenStemmer.stem("buying", nil)
-      "buying"
-
       iex> SimpleBayes.Trainer.TokenStemmer.stem("buying", false)
       "buying"
 
-      iex> SimpleBayes.Trainer.TokenStemmer.stem("buying", true)
+      iex> SimpleBayes.Trainer.TokenStemmer.stem("buying", &Stemmer.stem/1)
       "buy"
   """
-  def stem(word, true), do: Stemmer.stem(word)
-  def stem(word, _),    do: word
+  def stem(word, stemmer) when is_function(stemmer), do: stemmer.(word)
+  def stem(word, _), do: word
 end

mix.exs

@@ -17,16 +17,16 @@ defmodule SimpleBayes.Mixfile do
   end
 
   def application do
-    [applications: [:logger, :math, :decimal, :stemmer]]
+    [applications: [:logger, :math, :decimal]]
   end
 
   defp deps do
     [
       {:ex_doc,  ">= 0.0.0", only: :dev},
       {:faker,   ">= 0.0.0", only: :test},
+      {:stemmer, "~> 1.0.0", only: :test},
       {:math,    ">= 0.0.0"},
       {:decimal, ">= 0.0.0"},
-      {:stemmer, "~> 1.0.0"}
     ]
   end
 

mix.lock

@@ -2,5 +2,4 @@
   "earmark": {:hex, :earmark, "1.0.1", "2c2cd903bfdc3de3f189bd9a8d4569a075b88a8981ded9a0d95672f6e2b63141", [:mix], []},
   "ex_doc": {:hex, :ex_doc, "0.13.0", "aa2f8fe4c6136a2f7cfc0a7e06805f82530e91df00e2bff4b4362002b43ada65", [:mix], [{:earmark, "~> 1.0", [hex: :earmark, optional: false]}]},
   "math": {:hex, :math, "0.2.0", "e7d94733ef0f3a9051a7de58a8f30aa13bee4c7824020580a39dd19174f9cadc", [:mix], []},
-  "faker": {:hex, :faker, "0.6.0", "2d2ff0879d6b10fab5fb47eb2c1149b811e4af780a17022aa249deb3a7156d2b", [:mix], []},
-  "stemmer": {:hex, :stemmer, "1.0.0", "f39cc4a8ee63eb2b39daa3d43e4dafef42087e24c2f6fafea0be6e6ba2ecf862", [:mix], []}}
+  "faker": {:hex, :faker, "0.6.0", "2d2ff0879d6b10fab5fb47eb2c1149b811e4af780a17022aa249deb3a7156d2b", [:mix], []}}

test/simple_bayes/model/bernoulli_test.exs

@@ -90,7 +90,7 @@ defmodule SimpleBayes.BernoulliTest do
     end
 
     test "stemming" do
-      result = SimpleBayes.init(model: :bernoulli, stem: true)
+      result = SimpleBayes.init(model: :bernoulli, stem: &Stemmer.stem/1)
                |> SimpleBayes.train(:apple, "buying apple")
                |> SimpleBayes.train(:banana, "buy banana")
                |> SimpleBayes.classify("buy apple")

test/simple_bayes/model/binarized_multinomial_test.exs

@@ -77,7 +77,7 @@ defmodule SimpleBayes.BinarizedMultinomialTest do
     end
 
     test "stemming" do
-      result = SimpleBayes.init(model: :binarized_multinomial, stem: true)
+      result = SimpleBayes.init(model: :binarized_multinomial, stem: &Stemmer.stem/1)
                |> SimpleBayes.train(:apple, "buying apple")
                |> SimpleBayes.train(:banana, "buy banana")
                |> SimpleBayes.classify("buy apple")

test/simple_bayes/model/multinomial_test.exs

@@ -55,7 +55,7 @@ defmodule SimpleBayes.MultinomialTest do
     end
 
     test "stemming" do
-      result = SimpleBayes.init(stem: true)
+      result = SimpleBayes.init(stem: &Stemmer.stem/1)
                |> SimpleBayes.train(:apple, "buying apple")
                |> SimpleBayes.train(:banana, "buy banana")
                |> SimpleBayes.classify("buy apple")