Merge pull request #94 from maciejkula/add_flake8

Add PEP8 checks to circle.

Author: maciejkula

.flake8rc

@@ -1,2 +1,4 @@
 [flake8]
-max-line-length = 100
+ignore = I100
+max-line-length = 100
+exclude = doc/*

circle.yml

@@ -6,5 +6,7 @@ dependencies:
 test:
   pre:
     - pip install pytest
+    - pip install flake8
   override:
+    - flake8 --config .flake8rc
     - mv ~/lightfm/lightfm ~/lightfm/_lightfm && py.test -xv ~/lightfm/tests

lightfm/_lightfm_fast.py

@@ -1,13 +1,13 @@
 try:
     # Import OpenMP-enabled extension
-    from ._lightfm_fast_openmp import *
-    from ._lightfm_fast_openmp import __test_in_positives
+    from ._lightfm_fast_openmp import *  # NOQA
+    from ._lightfm_fast_openmp import __test_in_positives  # NOQA
 except ImportError:
     # Fall back on OpenMP-less extension
     import warnings
 
     warnings.warn('LightFM was compiled without OpenMP support. '
                   'Only a single thread will be used.')
 
-    from ._lightfm_fast_no_openmp import *
-    from ._lightfm_fast_no_openmp import __test_in_positives
+    from ._lightfm_fast_no_openmp import *  # NOQA
+    from ._lightfm_fast_no_openmp import __test_in_positives  # NOQA

lightfm/datasets/__init__.py

@@ -1,2 +1,2 @@
-from lightfm.datasets.movielens import fetch_movielens
-from lightfm.datasets.stackexchange import fetch_stackexchange
+from lightfm.datasets.movielens import fetch_movielens  # NOQA
+from lightfm.datasets.stackexchange import fetch_stackexchange  # NOQA

lightfm/datasets/movielens.py

@@ -1,12 +1,8 @@
-import array
 import itertools
-import os
 import zipfile
 
 import numpy as np
 
-import requests
-
 import scipy.sparse as sp
 
 from lightfm.datasets import _common
@@ -113,14 +109,16 @@ def fetch_movielens(data_home=None, indicator_features=True, genre_features=Fals
     """
     Fetch the `Movielens 100k dataset <http://grouplens.org/datasets/movielens/100k/>`_.
 
-    The dataset contains 100,000 interactions from 1000 users on 1700 movies, and is exhaustively described
-    in its `README <http://files.grouplens.org/datasets/movielens/ml-100k-README.txt>`_.
+    The dataset contains 100,000 interactions from 1000 users on 1700 movies,
+    and is exhaustively described in its
+    `README <http://files.grouplens.org/datasets/movielens/ml-100k-README.txt>`_.
 
     Parameters
     ----------
 
     data_home: path, optional
-        Path to the directory in which the downloaded data should be placed. Defaults to ``~/lightfm_data/``.
+        Path to the directory in which the downloaded data should be placed.
+        Defaults to ``~/lightfm_data/``.
     indicator_features: bool, optional
         Use an [n_users, n_users] identity matrix for item features. When True with genre_features,
         indicator and genre features are concatenated into a single feature matrix of shape

lightfm/datasets/stackexchange.py

@@ -97,8 +97,8 @@ def fetch_stackexchange(dataset, test_set_fraction=0.2,
                                      shape=data['features_shape'].flatten())
     tag_labels = data['labels']
 
-    test_cutoff_timestamp = np.sort(interactions.data)[len(interactions.data)
-                                                       * (1.0 - test_set_fraction)]
+    test_cutoff_timestamp = np.sort(interactions.data)[len(interactions.data) *
+                                                       (1.0 - test_set_fraction)]
     in_train = interactions.data < test_cutoff_timestamp
     in_test = np.logical_not(in_train)
 

lightfm/evaluation.py

@@ -51,8 +51,8 @@ def precision_at_k(model, test_interactions, train_interactions=None,
     -------
 
     np.array of shape [n_users with interactions or n_users,]
-         Numpy array containing precision@k scores for each user. If there are no interactions for a given
-         user the returned precision will be 0.
+         Numpy array containing precision@k scores for each user. If there are no interactions
+         for a given user the returned precision will be 0.
     """
 
     ranks = model.predict_rank(test_interactions,
@@ -146,8 +146,8 @@ def reciprocal_rank(model, test_interactions, train_interactions=None,
                     user_features=None, item_features=None,
                     preserve_rows=False, num_threads=1):
     """
-    Measure the reciprocal rank metric for a model: 1 / the rank of the highest ranked positive example.
-    A perfect score is 1.0.
+    Measure the reciprocal rank metric for a model: 1 / the rank of the highest
+    ranked positive example. A perfect score is 1.0.
 
     Parameters
     ----------
@@ -176,8 +176,8 @@ def reciprocal_rank(model, test_interactions, train_interactions=None,
     -------
 
     np.array of shape [n_users with interactions or n_users,]
-         Numpy array containing reciprocal rank scores for each user. If there are no interactions for a given
-         user the returned value will be 0.0.
+         Numpy array containing reciprocal rank scores for each user.
+         If there are no interactions for a given user the returned value will be 0.0.
     """
 
     ranks = model.predict_rank(test_interactions,

lightfm/lightfm.py

@@ -5,9 +5,8 @@
 import scipy.sparse as sp
 
 from ._lightfm_fast import (CSRMatrix, FastLightFM,
-                            fit_logistic, predict_lightfm,
-                            predict_ranks,
-                            fit_warp, fit_bpr, fit_warp_kos)
+                            fit_bpr, fit_logistic, fit_warp,
+                            fit_warp_kos, predict_lightfm, predict_ranks)
 
 
 __all__ = ['LightFM']
@@ -181,17 +180,17 @@ def _initialize(self, no_components, no_item_features, no_user_features):
         """
 
         # Initialise item features.
-        self.item_embeddings = ((self.rng.rand(no_item_features, no_components) - 0.5)
-                                / no_components).astype(np.float32)
+        self.item_embeddings = ((self.rng.rand(no_item_features, no_components) - 0.5) /
+                                no_components).astype(np.float32)
         self.item_embedding_gradients = np.zeros_like(self.item_embeddings)
         self.item_embedding_momentum = np.zeros_like(self.item_embeddings)
         self.item_biases = np.zeros(no_item_features, dtype=np.float32)
         self.item_bias_gradients = np.zeros_like(self.item_biases)
         self.item_bias_momentum = np.zeros_like(self.item_biases)
 
         # Initialise user features.
-        self.user_embeddings = ((self.rng.rand(no_user_features, no_components) - 0.5)
-                                / no_components).astype(np.float32)
+        self.user_embeddings = ((self.rng.rand(no_user_features, no_components) - 0.5) /
+                                no_components).astype(np.float32)
         self.user_embedding_gradients = np.zeros_like(self.user_embeddings)
         self.user_embedding_momentum = np.zeros_like(self.user_embeddings)
         self.user_biases = np.zeros(no_user_features, dtype=np.float32)
@@ -271,8 +270,7 @@ def _process_sample_weight(self, interactions, sample_weight):
                                  'matrices must be the same shape')
 
             if not (np.array_equal(interactions.row,
-                                   sample_weight.row)
-                    and
+                                   sample_weight.row) and
                     np.array_equal(interactions.col,
                                    sample_weight.col)):
                 raise ValueError('Sample weight and interaction matrix '
@@ -608,17 +606,18 @@ def predict_rank(self, test_interactions, train_interactions=None,
                      item_features=None, user_features=None, num_threads=1):
         """
         Predict the rank of selected interactions. Computes recommendation rankings across all items
-        for every user in interactions and calculates the rank of all non-zero entries in the recommendation
-        ranking, with 0 meaning the top of the list (most recommended) and n_items - 1 being the end of the
-        list (least recommended).
+        for every user in interactions and calculates the rank of all non-zero entries
+        in the recommendation ranking, with 0 meaning the top of the list (most recommended)
+        and n_items - 1 being the end of the list (least recommended).
 
         Arguments
         ---------
         test_interactions: np.float32 csr_matrix of shape [n_users, n_items]
              Non-zero entries denote the user-item pairs whose rank will be computed.
         train_interactions: np.float32 csr_matrix of shape [n_users, n_items], optional
-             Non-zero entries denote the user-item pairs which will be excluded from rank computation.
-             Use to exclude training set interactions from being scored and ranked for evaluation.
+             Non-zero entries denote the user-item pairs which will be excluded from
+             rank computation. Use to exclude training set interactions from being scored
+             and ranked for evaluation.
         item_ids: np.int32 array of shape [n_pairs,]
              an array containing the item ids for the user-item pairs for which
              a prediction is to be computed.
@@ -632,7 +631,7 @@ def predict_rank(self, test_interactions, train_interactions=None,
 
         Returns
         -------
-                                   
+
         np.float32 csr_matrix of shape [n_users, n_items]
             the [i, j]-th entry of the matrix will contain the rank of the j-th item
             in the sorted recommendations list for the i-th user. The degree

setup.py

@@ -1,5 +1,4 @@
 import os
-import platform
 import subprocess
 import sys
 import textwrap
@@ -75,7 +74,6 @@ def generate_pyx(self):
 
     def run(self):
 
-        import Cython
         from Cython.Build import cythonize
 
         self.generate_pyx()
@@ -129,7 +127,6 @@ def run_tests(self):
         sys.exit(errno)
 
 
-
 use_openmp = not (sys.platform.startswith('darwin') or sys.platform.startswith('win'))
 
 

tests/test_api.py

@@ -119,10 +119,10 @@ def test_not_enough_features_fails():
                            no_items),
                           dtype=np.int32)
 
-    user_features = sp.csr_matrix((no_users-1,
+    user_features = sp.csr_matrix((no_users - 1,
                                    no_features),
                                   dtype=np.int32)
-    item_features = sp.csr_matrix((no_items-1,
+    item_features = sp.csr_matrix((no_items - 1,
                                    no_features),
                                   dtype=np.int32)
     model = LightFM()
@@ -261,8 +261,8 @@ def test_predict_ranks():
     # in train in that row
     ranks = model.predict_rank(rank_input,
                                train_interactions=train).todense()
-    assert np.all(np.squeeze(np.array(ranks.max(axis=1)))
-                  == no_items - 1 - np.squeeze(np.array(train.getnnz(axis=1))))
+    assert np.all(np.squeeze(np.array(ranks.max(axis=1))) ==
+                  no_items - 1 - np.squeeze(np.array(train.getnnz(axis=1))))
 
     # Make sure invariants hold when there are ties
     model.user_embeddings = np.zeros_like(model.user_embeddings)