add LICENSE, code, img

LICENSE

@@ -0,0 +1,29 @@
+BSD 3-Clause License
+
+Copyright (c) 2017, babylon health
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+* Redistributions of source code must retain the above copyright notice, this
+  list of conditions and the following disclaimer.
+
+* Redistributions in binary form must reproduce the above copyright notice,
+  this list of conditions and the following disclaimer in the documentation
+  and/or other materials provided with the distribution.
+
+* Neither the name of the copyright holder nor the names of its
+  contributors may be used to endorse or promote products derived from
+  this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

fasttext.py

@@ -0,0 +1,66 @@
+#
+# Copyright (c) 2017-present, babylon health
+# All rights reserved.
+#
+# This source code is licensed under the BSD-style license found in the
+# LICENSE file in the root directory of this source tree.
+#
+
+import numpy as np
+
+
+class FastVector:
+    """
+    Minimal wrapper for fastvector embeddings.
+    ```
+    Usage:
+        $ model = FastVector(vector_file='/path/to/wiki.en.vec')
+        $ 'apple' in model
+        > TRUE
+        $ model['apple'].shape
+        > (300,)
+    ```
+    """
+
+    def __init__(self, vector_file='', transform=None):
+        """Read in word vectors in fasttext format"""
+        self.word2id = {}
+        print('reading word vectors from %s' % vector_file)
+        with open(vector_file, 'r') as f:
+            (self.n_words, self.n_dim) = \
+                (int(x) for x in f.readline().rstrip('\n').split(' '))
+            self.embed = np.zeros((self.n_words, self.n_dim))
+            for i, line in enumerate(f):
+                elems = line.rstrip('\n').split(' ')
+                self.word2id[elems[0]] = i
+                self.embed[i] = elems[1:self.n_dim+1]
+
+        if transform is not None:
+            print('Applying transformation to embedding')
+            self.apply_transform(transform)
+
+    def apply_transform(self, transform):
+        """
+        Apply the given transformation to the vector space
+
+        Right-multiplies given transform with embeddings E:
+            E = E * transform
+
+        Transform can either be a string with a filename to a
+        text file containing a ndarray (compat. with np.loadtxt)
+        or a numpy ndarray.
+        """
+        transmat = np.loadtxt(transform) if isinstance(transform, str) else transform
+        self.embed = np.matmul(self.embed, transmat)
+
+    @classmethod
+    def cosine_similarity(cls, vec_a, vec_b):
+        """Compute cosine similarity between vec_a and vec_b"""
+        return np.dot(vec_a, vec_b) / \
+            (np.linalg.norm(vec_a) * np.linalg.norm(vec_b))
+
+    def __contains__(self, key):
+        return key in self.word2id
+
+    def __getitem__(self, key):
+        return self.embed[self.word2id[key]]