Add JS variation of the BernoulliNB estimator

readme.md

@@ -121,7 +121,7 @@ Transpile trained [scikit-learn](https://github.com/scikit-learn/scikit-learn) e
             <td><a href="http://scikit-learn.org/stable/modules/generated/sklearn.naive_bayes.BernoulliNB.html#sklearn.naive_bayes.BernoulliNB">sklearn.naive_bayes.BernoulliNB</a></td>
             <td align="center"></td>
             <td align="center"><a href="examples/classifier/BernoulliNB/java/basics.ipynb">✓</a></td>
-            <td align="center"></td>
+            <td align="center">✓</td>
             <td align="center"></td>
             <td align="center"></td>
             <td align="center"></td>

sklearn_porter/classifier/BernoulliNB/__init__.py

@@ -22,6 +22,13 @@ class BernoulliNB(Classifier):
             'arr[][]':  '{type}[][] {name} = {{{values}}};',
             'indent':   '    ',
         },
+        'js': {
+            'type':     '{0}',
+            'arr':      '[{0}]',
+            'arr[]':    'var {name} = [{values}];',
+            'arr[][]':  'var {name} = [{values}];',
+            'indent':   '    ',
+        }
     }
     # @formatter:on
 
@@ -139,7 +146,7 @@ def create_method(self):
         :return out : string
             The built method as string.
         """
-        n_indents = 1 if self.target_language in ['java'] else 0
+        n_indents = 1 if self.target_language in ['java', 'js'] else 0
         temp_method = self.temp('method.predict', n_indents=n_indents,
                                 skipping=True)
         out = temp_method.format(**self.__dict__)

sklearn_porter/classifier/BernoulliNB/templates/js/class.txt

@@ -0,0 +1,13 @@
+var {class_name} = function() {{
+
+    {method}
+
+}};
+
+if (typeof process !== 'undefined' && typeof process.argv !== 'undefined') {{
+    if (process.argv.length - 2 == {n_features}) {{
+        var argv = process.argv.slice(2);
+        var prediction = new {class_name}().{method_name}(argv);
+        console.log(prediction);
+    }}
+}}

sklearn_porter/classifier/BernoulliNB/templates/js/method.predict.txt

@@ -0,0 +1,33 @@
+this.{method_name} = function(atts) {{
+
+    {priors}
+    {neg_probs}
+    {del_probs}
+
+    var jll = new Array({n_classes});
+    for (var i = 0; i < {n_classes}; i++) {{
+        var sum = 0.;
+        for (var j = 0; j < {n_features}; j++) {{
+            sum += atts[i] * delProbs[j][i];
+        }}
+        jll[i] = sum;
+    }}
+    for (var i = 0; i < {n_classes}; i++) {{
+        var sum = 0.;
+        for (var j = 0; j < {n_features}; j++) {{
+            sum += negProbs[i][j];
+        }}
+        jll[i] += priors[i] + sum;
+    }}
+
+    var highestLikeli = -1;
+    var classIndex = -1;
+    for (var i = 0; i < {n_classes}; i++) {{
+        if (jll[i] > highestLikeli) {{
+            highestLikeli = jll[i];
+            classIndex = i;
+        }}
+    }}
+    return classIndex;
+}};
+

tests/classifier/BernoulliNB/BernoulliNBJSTest.py

@@ -0,0 +1,38 @@
+# -*- coding: utf-8 -*-
+
+import unittest
+
+from sklearn.naive_bayes import BernoulliNB
+
+from ..Classifier import Classifier
+from ...language.JavaScript import JavaScript
+
+
+class BernoulliNBJSTest(JavaScript, Classifier, unittest.TestCase):
+
+    def setUp(self):
+        super(BernoulliNBJSTest, self).setUp()
+        self.mdl = BernoulliNB()
+
+    def tearDown(self):
+        super(BernoulliNBJSTest, self).tearDown()
+
+    @unittest.skip('BernoulliNB is just suitable for discrete data.')
+    def test_random_features_w_iris_data(self):
+        pass
+
+    @unittest.skip('BernoulliNB is just suitable for discrete data.')
+    def test_existing_features_w_binary_data(self):
+        pass
+
+    @unittest.skip('BernoulliNB is just suitable for discrete data.')
+    def test_random_features_w_binary_data(self):
+        pass
+
+    @unittest.skip('BernoulliNB is just suitable for discrete data.')
+    def test_random_features_w_digits_data(self):
+        pass
+
+    @unittest.skip('BernoulliNB is just suitable for discrete data.')
+    def test_existing_features_w_digits_data(self):
+        pass