diff --git a/examples/cumbo/plot_cumbo_2_views_2_classes.py b/examples/cumbo/plot_cumbo_2_views_2_classes.py
index c669f6919c71f964891940bdfa3c1276e7342992..77d0588aafa3392199880800c9ad33cba959f094 100644
--- a/examples/cumbo/plot_cumbo_2_views_2_classes.py
+++ b/examples/cumbo/plot_cumbo_2_views_2_classes.py
@@ -97,10 +97,10 @@ for ind_view in range(2):
label='Class {} ({})'.format(ind_class, label))
ax.legend()
-print('\nThe second figure displays the classification results for the '
- 'sub-classifiers\non the learning sample data.\n')
-
-styles = ('.b', '.r')
+# print('\nThe second figure displays the classification results for the '
+# 'sub-classifiers\non the learning sample data.\n')
+#
+# styles = ('.b', '.r')
# fig = plt.figure(figsize=(12., 7.))
# fig.suptitle('Classification results on the learning data for the '
# 'sub-classifiers', size=16)
diff --git a/examples/mvml/plot_mvml_.py b/examples/mvml/plot_mvml_.py
index 3120cb00eaddb579972db21be7065746dbff472b..8989230d812287c5c4fefde4dcc9aecbfd2d8a48 100644
--- a/examples/mvml/plot_mvml_.py
+++ b/examples/mvml/plot_mvml_.py
@@ -157,7 +157,7 @@ plt.subplot(342)
plt.scatter(X1[n_tr:n_tot, 0], X1[n_tr:n_tot, 1], c=Y[n_tr:n_tot])
plt.title("orig. view 2")
#
-pred1[np.where(pred1[:, 0] != Y[n_tr:n_tot])] = 0
+pred1[np.where(pred1[:] != Y[n_tr:n_tot])] = 0
pred1 = pred1.reshape((pred1.shape[0]))
plt.subplot(343)
plt.scatter(X0[n_tr:n_tot, 0], X0[n_tr:n_tot, 1], c=pred1)
@@ -166,7 +166,7 @@ plt.subplot(344)
plt.scatter(X1[n_tr:n_tot, 0], X1[n_tr:n_tot, 1], c=pred1)
plt.title("MVML view 2")
#
-pred2[np.where(pred2[:, 0] != Y[n_tr:n_tot])] = 0
+pred2[np.where(pred2[:] != Y[n_tr:n_tot])] = 0
pred2 = pred2.reshape((pred2.shape[0]))
plt.subplot(345)
plt.scatter(X0[n_tr:n_tot, 0], X0[n_tr:n_tot, 1], c=pred2)
@@ -175,7 +175,7 @@ plt.subplot(346)
plt.scatter(X1[n_tr:n_tot, 0], X1[n_tr:n_tot, 1], c=pred2)
plt.title("MVMLsparse view 2")
#
-pred3[np.where(pred3[:, 0] != Y[n_tr:n_tot])] = 0
+pred3[np.where(pred3[:] != Y[n_tr:n_tot])] = 0
pred3 = pred3.reshape((pred3.shape[0]))
#
plt.subplot(347)
@@ -185,7 +185,7 @@ plt.subplot(348)
plt.scatter(X1[n_tr:n_tot, 0], X1[n_tr:n_tot, 1], c=pred3)
plt.title("MVML_Cov view 2")
#
-pred4[np.where(pred4[:, 0] != Y[n_tr:n_tot])] = 0
+pred4[np.where(pred4[:] != Y[n_tr:n_tot])] = 0
pred4 = pred4.reshape((pred4.shape[0]))
plt.subplot(349)
plt.scatter(X0[n_tr:n_tot, 0], X0[n_tr:n_tot, 1], c=pred4)
@@ -194,7 +194,7 @@ plt.subplot(3,4,10)
plt.scatter(X1[n_tr:n_tot, 0], X1[n_tr:n_tot, 1], c=pred4)
plt.title("MVML_I view 2")
#
-pred5[np.where(pred5[:, 0] != Y[n_tr:n_tot])] = 0
+pred5[np.where(pred5[:] != Y[n_tr:n_tot])] = 0
pred5 = pred5.reshape((pred5.shape[0]))
plt.subplot(3,4,11)
plt.scatter(X0[n_tr:n_tot, 0], X0[n_tr:n_tot, 1], c=pred5)
diff --git a/multimodal/boosting/boost.py b/multimodal/boosting/boost.py
index 706350318258c25d35a6afb5aafcd57811663140..e6536e27acf84fef0f11e05678eee74388bc9b00 100644
--- a/multimodal/boosting/boost.py
+++ b/multimodal/boosting/boost.py
@@ -50,12 +50,12 @@ class UBoosting(metaclass=ABCMeta):
def _global_X_transform(self, X, views_ind=None):
X_ = None
- if isinstance(X, sp.spmatrix):
+ if isinstance(X, MultiModalData):
+ X_ = X
+ elif isinstance(X, sp.spmatrix):
X_ = MultiModalSparseArray(X, views_ind)
else:
X_ = MultiModalArray(X, views_ind)
- if isinstance(X, MultiModalData):
- X_ = X
if not isinstance(X_, MultiModalData):
try:
X_ = np.asarray(X)
diff --git a/multimodal/boosting/cumbo.py b/multimodal/boosting/cumbo.py
index 0d8d42c5cade316febd1f6b3b3feed09e35cb7fa..b498064cc18f703c9afc7d50811c947f87585651 100644
--- a/multimodal/boosting/cumbo.py
+++ b/multimodal/boosting/cumbo.py
@@ -99,15 +99,15 @@ class MuCumboClassifier(BaseEnsemble, ClassifierMixin, UBoosting):
>>> from sklearn.tree import DecisionTreeClassifier
>>> base_estimator = DecisionTreeClassifier(max_depth=2)
- >>> clf = MuCumboClassifier(base_estimator=base_estimator, random_state=0)
+ >>> clf = MuCumboClassifier(base_estimator=base_estimator, random_state=1)
>>> clf.fit(X, y, views_ind) # doctest: +NORMALIZE_WHITESPACE
MuCumboClassifier(base_estimator=DecisionTreeClassifier(class_weight=None, criterion='gini', max_depth=2,
- max_features=None, max_leaf_nodes=None,
- min_impurity_decrease=0.0, min_impurity_split=None,
- min_samples_leaf=1, min_samples_split=2,
- min_weight_fraction_leaf=0.0, presort=False, random_state=None,
- splitter='best'),
- n_estimators=50, random_state=0)
+ max_features=None, max_leaf_nodes=None,
+ min_impurity_decrease=0.0, min_impurity_split=None,
+ min_samples_leaf=1, min_samples_split=2,
+ min_weight_fraction_leaf=0.0, presort=False, random_state=None,
+ splitter='best'),
+ n_estimators=50, random_state=1)
>>> print(clf.predict([[ 5., 3., 1., 1.]]))
[0]
diff --git a/multimodal/datasets/data_sample.py b/multimodal/datasets/data_sample.py
index eff5b3dfbe6553e5053cc0ea985752683297eb5f..29fd0e60abdee29e2070bd80c71c1a386333e0e0 100644
--- a/multimodal/datasets/data_sample.py
+++ b/multimodal/datasets/data_sample.py
@@ -128,39 +128,39 @@ class MultiModalData(metaclass=ABCMeta):
class MultiModalSparseInfo():
- def __init__(self, data, view_ind=None):
+ def __init__(self, data, views_ind=None):
"""Constructor of Metriclearn_array"""
shapes_int = []
index = 0
new_data = np.ndarray([])
n_views = data.size
thekeys = None
- # view_ind_self = None
+ # views_ind_self = None
view_mode = 'slices'
if (sp.issparse(data)) and data.ndim > 1:
- if view_ind is not None:
+ if views_ind is not None:
try:
- view_ind = np.asarray(view_ind)
+ views_ind = np.asarray(views_ind)
except :
raise TypeError("n_views should be list or nparray")
- elif view_ind is None:
+ elif views_ind is None:
if data.shape[1] > 1:
- view_ind = np.array([0, data.shape[1]//2, data.shape[1]])
+ views_ind = np.array([0, data.shape[1]//2, data.shape[1]])
else:
- view_ind = np.array([0, data.shape[1]])
+ views_ind = np.array([0, data.shape[1]])
new_data = data
- # view_ind_self = view_ind
- view_ind, n_views, view_mode = self._first_validate_views_ind(view_ind,
+ # views_ind_self = views_ind
+ views_ind, n_views, view_mode = self._first_validate_views_ind(views_ind,
data.shape[1])
- if view_ind.ndim == 1 and view_mode.startswith("slicing"):
- shapes_int = [in2 - in1 for in1, in2 in zip(view_ind, view_ind[1:])]
+ if views_ind.ndim == 1 and view_mode.startswith("slicing"):
+ shapes_int = [in2 - in1 for in1, in2 in zip(views_ind, views_ind[1:])]
if data.shape[0] < 1 or data.shape[1] < 1:
raise ValueError("input data shouldbe not empty")
self.view_mode_ = view_mode
- self.views_ind = view_ind
+ self.views_ind = views_ind
self.shapes_int = shapes_int
self.n_views = n_views
@@ -201,7 +201,7 @@ class MultiModalSparseArray(sp.csr_matrix, sp.csc_matrix, MultiModalSparseInfo,
Attributes
----------
- view_ind : list of views' indice (may be None)
+ views_ind : list of views' indice (may be None)
n_views : int number of views
@@ -274,7 +274,7 @@ class MultiModalArray(np.ndarray, MultiModalData):
Attributes
----------
- view_ind : list of views' indice (may be None)
+ views_ind : list of views' indice (may be None)
n_views : int number of views
@@ -300,32 +300,31 @@ class MultiModalArray(np.ndarray, MultiModalData):
"""
- def __new__(cls, data, view_ind=None):
+ def __new__(cls, data, views_ind=None):
"""Constructor of MultiModalArray_array"""
shapes_int = []
index = 0
new_data = np.ndarray([])
n_views = 1
thekeys = None
- # view_ind_self = None
view_mode = 'slices'
if isinstance(data, dict) and not isinstance(data, sp.dok_matrix):
n_views = len(data)
- view_ind = [0]
+ views_ind = [0]
for key, dat_values in data.items():
dat_values = np.asarray(dat_values)
if dat_values.ndim < 2:
dat_values = dat_values.reshape(1, dat_values.shape[0])
new_data = cls._populate_new_data(index, dat_values, new_data)
shapes_int.append(dat_values.shape[1])
- view_ind.append(dat_values.shape[1] + view_ind[index])
+ views_ind.append(dat_values.shape[1] + views_ind[index])
index += 1
thekeys = data.keys()
if new_data.ndim < 2 :
raise ValueError('Reshape your data')
if new_data.ndim > 1 and (new_data.shape == (1, 1) or new_data.shape == ()):
raise ValueError('Reshape your data')
- elif isinstance(data, np.ndarray) and view_ind is None and data.ndim == 1:
+ elif isinstance(data, np.ndarray) and views_ind is None and data.ndim == 1:
try:
dat0 = np.array(data[0])
except Exception:
@@ -334,12 +333,12 @@ class MultiModalArray(np.ndarray, MultiModalData):
if dat0.ndim < 2:
data = data[np.newaxis, ...]
if data.shape[1] > 1:
- view_ind = np.array([0, data.shape[1]//2, data.shape[1]])
+ views_ind = np.array([0, data.shape[1]//2, data.shape[1]])
else:
- view_ind = np.array([0, data.shape[1]])
+ views_ind = np.array([0, data.shape[1]])
new_data = data
else:
- new_data, shapes_int, view_ind = cls._for_data(cls, data)
+ new_data, shapes_int, views_ind = cls._for_data(cls, data)
n_views = data.shape[0]
elif (isinstance(data, np.ndarray) ) and data.ndim > 1:
try:
@@ -347,41 +346,41 @@ class MultiModalArray(np.ndarray, MultiModalData):
except:
raise TypeError("input format is not supported")
- if view_ind is not None:
+ if views_ind is not None:
try:
- view_ind = np.asarray(view_ind)
+ views_ind = np.asarray(views_ind)
except :
raise TypeError("n_views should be list or nparray")
- elif view_ind is None:
+ elif views_ind is None:
if data.shape[1] > 1:
- view_ind = np.array([0, data.shape[1]//2, data.shape[1]])
+ views_ind = np.array([0, data.shape[1]//2, data.shape[1]])
else:
- view_ind = np.array([0, data.shape[1]])
+ views_ind = np.array([0, data.shape[1]])
new_data = data
else:
try:
new_data = np.asarray(data)
- if view_ind is None:
- view_ind = np.array([0, new_data.shape[1]])
+ if views_ind is None:
+ views_ind = np.array([0, new_data.shape[1]])
except Exception as e:
raise ValueError('Reshape your data')
if new_data.ndim < 2 :
raise ValueError('Reshape your data')
if new_data.ndim > 1 and (new_data.shape == (1, 1) or new_data.shape == ()):
raise ValueError('Reshape your data')
- if view_ind.ndim < 2 and new_data.ndim <2 and view_ind[-1] > new_data.shape[1]:
+ if views_ind.ndim < 2 and new_data.ndim <2 and views_ind[-1] > new_data.shape[1]:
raise ValueError('Reshape your data')
- # view_ind_self = view_ind
+ # views_ind_self = views_ind
# if new_data.shape[1] < 1:
# msg = ("%d feature\(s\) \\(shape=\%s\) while a minimum of \\d* "
# "is required.") % (new_data.shape[1], str(new_data.shape))
# # "%d feature\(s\) \(shape=\(%d, %d\)\) while a minimum of \d* is required." % (new_data.shape[1], new_data.shape[0], new_data.shape[1])
# raise ValueError(msg)
- view_ind, n_views, view_mode = cls._first_validate_views_ind(view_ind,
+ views_ind, n_views, view_mode = cls._first_validate_views_ind(views_ind,
new_data.shape[1])
- if view_ind.ndim == 1 and view_mode.startswith("slices"):
- shapes_int = [in2 - in1 for in1, in2 in zip(view_ind, view_ind[1:])]
+ if views_ind.ndim == 1 and view_mode.startswith("slices"):
+ shapes_int = [in2 - in1 for in1, in2 in zip(views_ind, views_ind[1:])]
# obj = ma.MaskedArray.__new(new_data) # new_data.view() a.MaskedArray(new_data, mask=new_data.mask).view(cls)
# bj = super(Metriclearn_array, cls).__new__(cls, new_data.data, new_data.mask)
@@ -393,7 +392,7 @@ class MultiModalArray(np.ndarray, MultiModalData):
else:
obj = np.recarray.__new__(cls, shape=(0, 0), dtype=np.float)
obj.view_mode_ = view_mode
- obj.views_ind = view_ind
+ obj.views_ind = views_ind
obj.shapes_int = shapes_int
obj.n_views = n_views
# obj.keys = thekeys
@@ -403,8 +402,8 @@ class MultiModalArray(np.ndarray, MultiModalData):
def _for_data(cls, data):
n_views = data.shape[0]
index = 0
- view_ind = np.empty(n_views + 1, dtype=np.int)
- view_ind[0] = 0
+ views_ind = np.empty(n_views + 1, dtype=np.int)
+ views_ind[0] = 0
shapes_int = []
new_data = np.ndarray([])
for dat_values in data:
@@ -413,10 +412,10 @@ class MultiModalArray(np.ndarray, MultiModalData):
except Exception:
raise TypeError("input format is not supported")
new_data = cls._populate_new_data(index, dat_values, new_data)
- view_ind[index + 1] = dat_values.shape[1] + view_ind[index]
+ views_ind[index + 1] = dat_values.shape[1] + views_ind[index]
shapes_int.append(dat_values.shape[1])
index += 1
- return new_data, shapes_int, view_ind
+ return new_data, shapes_int, views_ind
@staticmethod
def _populate_new_data(index, dat_values, new_data):
diff --git a/multimodal/kernels/lpMKL.py b/multimodal/kernels/lpMKL.py
index cb051c03540f654cd5fc49e8edb70a602a86661a..ca61b8cbf283c62192c26d30ee26abed419dd192 100644
--- a/multimodal/kernels/lpMKL.py
+++ b/multimodal/kernels/lpMKL.py
@@ -193,7 +193,6 @@ class MKL(BaseEstimator, ClassifierMixin, MKernel):
# ft2[v] = weights_old[v] * np.dot(np.transpose(C), np.dot(data.kernel_dict[v], C))
# calculate the sum for downstairs
# print(weights_old)
- # print(ft2)
# print(ft2 ** (p / (p + 1.0)))
downstairs = np.sum(ft2 ** (p / (p + 1.0))) ** (1.0 / p)
# and then the gammas
diff --git a/multimodal/kernels/mvml.py b/multimodal/kernels/mvml.py
index e1b22f27c7f3ebfd51892aad234d031198c09d78..edcf0935d2f4928e12887ed5a07b70064b3010e7 100644
--- a/multimodal/kernels/mvml.py
+++ b/multimodal/kernels/mvml.py
@@ -99,7 +99,41 @@ class MVML(MKernel, BaseEstimator, ClassifierMixin):
regression_ : if the classifier is used as regression (default : False)
-
+ Examples
+ --------
+ >>> from multimodal.kernels.mvml import MVML
+ >>> from sklearn.datasets import load_iris
+ >>> X, y = load_iris(return_X_y=True)
+ >>> views_ind = [0, 2, 4] # view 0: sepal data, view 1: petal data
+ >>> clf = MVML()
+ clf.get_params()
+ {'eta': 1, 'kernel': 'linear', 'kernel_params': None, 'learn_A': 1, 'learn_w': 0, 'lmbda': 0.1, 'n_loops': 6, 'nystrom_param': 1.0, 'precision': 0.0001}
+ >>> clf.fit(X, y, views_ind) # doctest: +NORMALIZE_WHITESPACE
+ MumboClassifier(base_estimator=None, best_view_mode='edge',
+ n_estimators=50, random_state=0)
+ >>> print(clf.predict([[ 5., 3., 1., 1.]]))
+ [1]
+ >>> views_ind = [[0, 2], [1, 3]] # view 0: length data, view 1: width data
+ >>> clf = MumboClassifier(random_state=0)
+ >>> clf.fit(X, y, views_ind) # doctest: +NORMALIZE_WHITESPACE
+ MumboClassifier(base_estimator=None, best_view_mode='edge',
+ n_estimators=50, random_state=0)
+ >>> print(clf.predict([[ 5., 3., 1., 1.]]))
+ [1]
+
+ >>> from sklearn.tree import DecisionTreeClassifier
+ >>> base_estimator = DecisionTreeClassifier(max_depth=2)
+ >>> clf = MumboClassifier(base_estimator=base_estimator, random_state=0)
+ >>> clf.fit(X, y, views_ind) # doctest: +NORMALIZE_WHITESPACE
+ MumboClassifier(base_estimator=DecisionTreeClassifier(class_weight=None,
+ criterion='gini', max_depth=2, max_features=None,
+ max_leaf_nodes=None, min_impurity_decrease=0.0,
+ min_impurity_split=None, min_samples_leaf=1, min_samples_split=2,
+ min_weight_fraction_leaf=0.0, presort=False, random_state=None,
+ splitter='best'),
+ best_view_mode='edge', n_estimators=50, random_state=0)
+ >>> print(clf.predict([[ 5., 3., 1., 1.]]))
+ [1]
"""
# r_cond = 10-30
def __init__(self, lmbda=0.1, eta=1, nystrom_param=1.0, kernel="linear",
diff --git a/multimodal/tests/test_cumbo.py b/multimodal/tests/test_cumbo.py
index 49ec85c2d87b110c04d283e94b8a3d180d0790dc..f028e26fe46e93da31d6fe7c5bef8849804f5534 100644
--- a/multimodal/tests/test_cumbo.py
+++ b/multimodal/tests/test_cumbo.py
@@ -866,9 +866,15 @@ class TestMuCumboClassifier(unittest.TestCase):
'base_estimator__max_depth': (1, 2)}
clf = GridSearchCV(mumbo, parameters)
clf.fit(self.iris.data, self.iris.target, views_ind=self.iris.views_ind)
- #
- #
- # def test_pickle():
+ self.assertEqual(clf.best_params_,{'base_estimator__max_depth': 2, 'n_estimators': 2})
+
+ multimodal_data = MultiModalArray(self.iris.data, views_ind=self.iris.views_ind)
+ clf = GridSearchCV(mumbo, parameters)
+ clf.fit(multimodal_data, self.iris.target)
+
+ self.assertEqual(clf.best_params_, {'base_estimator__max_depth': 2, 'n_estimators': 2})
+
+ # def test_pick le():
# np.random.seed(seed)
#
# # Check pickability.
diff --git a/multimodal/tests/test_mumbo.py b/multimodal/tests/test_mumbo.py
index f22f5ba72ea6d747c9f80be486830198c3f2ddb4..7082462eea6ffe34de2960228a1659a93089e882 100644
--- a/multimodal/tests/test_mumbo.py
+++ b/multimodal/tests/test_mumbo.py
@@ -786,6 +786,7 @@ class TestMuCumboClassifier(unittest.TestCase):
clf = GridSearchCV(mumbo, parameters)
clf.fit(self.iris.data, self.iris.target, views_ind=self.iris.views_ind)
+
def test_pickle(self):
seed = 7
np.random.seed(seed)