Merge d398793 into d8bc3d8

.travis.yml

@@ -1,19 +1,15 @@
+language: python
+
 sudo: false
 
+python:
+  - 2.7
+  - 3.6
+
 env:
   global:
     - CONDA_DEPS="pip pytest numpy pandas scipy matplotlib seaborn"
-
-matrix:
-  include:
-    # - os: linux
-    #   env:
-    #     - PYTHON_VERSION=3.5
-    #     - MINICONDA_URL="https://repo.continuum.io/miniconda/Miniconda-latest-Linux-x86_64.sh"
-    - os: linux
-      env:
-         - PYTHON_VERSION=2.7
-         - MINICONDA_URL="https://repo.continuum.io/miniconda/Miniconda-latest-Linux-x86_64.sh"
+    - MINICONDA_URL="https://repo.continuum.io/miniconda/Miniconda-latest-Linux-x86_64.sh"
 
 before_install:
 - export MINICONDA=$HOME/miniconda
@@ -26,7 +22,7 @@ before_install:
 - conda update conda
 - conda info -a
 - conda config --add channels conda-forge
-- conda install python=$PYTHON_VERSION $CONDA_DEPS
+- conda install python=$TRAVIS_PYTHON_VERSION $CONDA_DEPS
 
 install:
 - pip install -r requirements.txt

emotioncf/tests/test_core.py

@@ -8,15 +8,15 @@
 def simulate_data(data_type = 'data_long'):
     i = 100
     s = 50
-    rat = np.random.rand(s,i)*50
-    for x in np.arange(0,rat.shape[1],5):
-        rat[0:s/2,x] = rat[0:s/2,x] + x
-    for x in np.arange(0,rat.shape[1],3):
-        rat[(s/2):s,x] = rat[(s/2):s,x] + x
-    rat[(s/2):s] = rat[(s/2):s,::-1]
+    rat = np.random.rand(s, i)*50
+    for x in np.arange(0, rat.shape[1], 5):
+        rat[0:int(s/2), x] = rat[0:int(s/2), x] + x
+    for x in np.arange(0, rat.shape[1], 3):
+        rat[int(s/2):s, x] = rat[int(s/2):s, x] + x
+    rat[int(s/2):s] = rat[int(s/2):s, ::-1]
     rat = pd.DataFrame(rat)
     if data_type is 'data_long':
-        out = pd.DataFrame(columns=['Subject','Item','Rating'])
+        out = pd.DataFrame(columns=['Subject', 'Item', 'Rating'])
         for row in rat.iterrows():
             sub = pd.DataFrame(columns=out.columns)
             sub['Rating'] = row[1]
@@ -28,14 +28,14 @@ def simulate_data(data_type = 'data_long'):
         return rat
 
 def basecf_method_test(cf=None, data=None):
-    assert cf.train_mask.shape == (50,100)
-    assert cf.predicted_ratings.shape == (50,100)
+    assert cf.train_mask.shape == (50, 100)
+    assert cf.predicted_ratings.shape == (50, 100)
     mse = cf.get_mse(data=data)
     r = cf.get_corr(data=data)
     sub_r = cf.get_sub_corr(data=data)
     sub_mse = cf.get_sub_mse(data=data)
-    assert isinstance(mse,float)
-    assert isinstance(r,float)
+    assert isinstance(mse, float)
+    assert isinstance(r, float)
     assert isinstance(sub_r,np.ndarray)
     assert len(sub_r) == cf.ratings.shape[0]
     assert isinstance(sub_mse,np.ndarray)
@@ -47,7 +47,7 @@ def basecf_method_test(cf=None, data=None):
     print(('mse: %s') % mse)
     print(('r: %s') % r)
     print(('mean sub r: %s') % np.mean(sub_r))
-    
+
     df = cf.to_long_df()
     assert isinstance(df,pd.DataFrame)
     if cf.is_predict:
@@ -57,18 +57,18 @@ def basecf_method_test(cf=None, data=None):
         assert df.shape[0] == cf.ratings.shape[0]*cf.ratings.shape[1]*2
     if cf.is_mask:
         assert 'Mask' in df.columns
-    
+
 def basecf_method_all_tests(cf=None):
     cf.plot_predictions()
     basecf_method_test(cf=cf, data='all')
     basecf_method_test(cf=cf, data='train')
     basecf_method_test(cf=cf, data='test')
-    
+
 def test_create_sub_by_item_matrix():
     rating = create_sub_by_item_matrix(simulate_data(data_type='data_long'))
     assert isinstance(rating,pd.DataFrame)
     assert rating.shape == (50,100)
-    
+
 def test_cf_mean():
     cf = Mean(simulate_data(data_type='data_wide'))
     cf.fit()
@@ -81,64 +81,64 @@ def test_cf_mean():
     cf.fit(dilate_ts_n_samples=2)
     cf.predict()
     basecf_method_all_tests(cf=cf)
-    
-    
+
+
 def test_cf_knn():
     cf = KNN(simulate_data(data_type='data_wide'))
     cf.fit(metric='correlation')
     cf.predict()
-    
+
     cf.split_train_test(n_train_items=50)
     cf.fit()
     cf.predict()
     basecf_method_all_tests(cf=cf)
-    
+
     cf.fit(metric='correlation')
     cf.predict(k=10)
     basecf_method_test(cf=cf, data='all')
-    
+
     cf.fit()
     cf.predict(k=10)
     basecf_method_all_tests(cf=cf)
 
-    cf.fit(dilate_ts_n_samples=2,metric='correlation')
+    cf.fit(dilate_ts_n_samples=2, metric='correlation')
     cf.predict()
     basecf_method_all_tests(cf=cf)
-    
+
     cf.fit(metric='cosine')
     cf.predict(k=10)
     basecf_method_test(cf=cf, data='all')
-    
+
     cf.fit()
     cf.predict(k=10)
     basecf_method_all_tests(cf=cf)
 
-    cf.fit(dilate_ts_n_samples=2,metric='cosine')
+    cf.fit(dilate_ts_n_samples=2, metric='cosine')
     cf.predict()
     basecf_method_all_tests(cf=cf)
-    
+
 def test_cf_knn_dil():
     cf = KNN(simulate_data(data_type='data_wide'))
     cf.split_train_test(n_train_items=20)
-    cf.fit(dilate_ts_n_samples=2,metric='correlation')
+    cf.fit(dilate_ts_n_samples=2, metric='correlation')
     cf.predict()
     basecf_method_all_tests(cf=cf)
 
-    cf.fit(dilate_ts_n_samples=2,metric='cosine')
+    cf.fit(dilate_ts_n_samples=2, metric='cosine')
     cf.predict()
     basecf_method_all_tests(cf=cf)
-    
+
 def test_cf_nnmf_multiplicative():
     cf = NNMF_multiplicative(simulate_data(data_type='data_wide'))
     cf.fit()
     cf.predict()
     cf.split_train_test(n_train_items=50)
     cf.fit()
     basecf_method_all_tests(cf=cf)
-    
+
     cf.fit(dilate_ts_n_samples=2)
     cf.predict()
-    basecf_method_all_tests(cf=cf)  
+    basecf_method_all_tests(cf=cf)
 
 def test_cf_nnmf_sgd():
     cf = NNMF_sgd(simulate_data(data_type='data_wide'))
@@ -149,7 +149,7 @@ def test_cf_nnmf_sgd():
            item_bias_reg=0,
            learning_rate=.001)
     cf.predict()
-    
+
     cf.split_train_test(n_train_items=50)
     cf.fit(n_iterations = 20,
            user_fact_reg=0,
@@ -158,7 +158,7 @@ def test_cf_nnmf_sgd():
            item_bias_reg=0,
            learning_rate=.001)
     basecf_method_all_tests(cf=cf)
-    
+
     cf.fit(n_iterations = 20,
            user_fact_reg=0,
            item_fact_reg=0,
@@ -167,24 +167,23 @@ def test_cf_nnmf_sgd():
            learning_rate=.001,
            dilate_ts_n_samples=2)
     cf.predict()
-    basecf_method_all_tests(cf=cf)  
-    
+    basecf_method_all_tests(cf=cf)
+
 def test_downsample():
     cf = Mean(simulate_data(data_type = 'data_wide'))
-    cf.downsample(sampling_freq=10,target=2, target_type='samples')
+    cf.downsample(sampling_freq=10, target=2, target_type='samples')
     assert cf.ratings.shape == (50,50)
     cf = Mean(simulate_data(data_type = 'data_wide'))
-    cf.downsample(sampling_freq=10,target=5, target_type='hz')
-    assert cf.ratings.shape == (50,50)
+    cf.downsample(sampling_freq=10, target=5, target_type='hz')
+    assert cf.ratings.shape == (50, 50)
     cf = Mean(simulate_data(data_type = 'data_wide'))
-    cf.downsample(sampling_freq=10,target=2, target_type='seconds')
-    assert cf.ratings.shape == (50,5)
+    cf.downsample(sampling_freq=10, target=2, target_type='seconds')
+    assert cf.ratings.shape == (50, 5)
     cf = Mean(simulate_data(data_type = 'data_wide'))
     cf.split_train_test(n_train_items=20)
     cf.fit()
     cf.predict()
     cf.downsample(sampling_freq=10,target=2, target_type='samples')
-    assert cf.ratings.shape == (50,50)
-    assert cf.train_mask.shape == (50,50)
-    assert cf.predicted_ratings.shape == (50,50) 
-
+    assert cf.ratings.shape == (50, 50)
+    assert cf.train_mask.shape == (50, 50)
+    assert cf.predicted_ratings.shape == (50, 50)