def test_rest(x, y):
print('Random under-sampling')
US = UnderSampler(indices_support=indices_support, verbose=verbose)
usx, usy, idx_tmp = US.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
print('Tomek links')
TL = TomekLinks(verbose=verbose)
tlx, tly = TL.fit_transform(x, y)
print('Clustering centroids')
CC = ClusterCentroids(verbose=verbose)
ccx, ccy = CC.fit_transform(x, y)
print('NearMiss-1')
NM1 = NearMiss(version=1, indices_support=indices_support, verbose=verbose)
nm1x, nm1y, idx_tmp = NM1.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
print('NearMiss-2')
NM2 = NearMiss(version=2, indices_support=indices_support, verbose=verbose)
nm2x, nm2y, idx_tmp = NM2.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
print('NearMiss-3')
NM3 = NearMiss(version=3, indices_support=indices_support, verbose=verbose)
nm3x, nm3y, idx_tmp = NM3.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
print('Neighboorhood Cleaning Rule')
NCR = NeighbourhoodCleaningRule(indices_support=indices_support, verbose=verbose)
ncrx, ncry, idx_tmp = NCR.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
print('Random over-sampling')
OS = OverSampler(verbose=verbose)
ox, oy = OS.fit_transform(x, y)
print('SMOTE Tomek links')
STK = SMOTETomek(verbose=verbose)
stkx, stky = STK.fit_transform(x, y)
print('SMOTE ENN')
SENN = SMOTEENN(verbose=verbose)
sennx, senny = SENN.fit_transform(x, y)
print('EasyEnsemble')
EE = EasyEnsemble(verbose=verbose)
eex, eey = EE.fit_transform(x, y)
def nearmiss_undersampling(X, y, version):
"""
Perform NearMiss undersampling
Keyword arguments:
X -- The feature vectors
y -- The target classes
"""
if verbose:
print '\nUndersampling with NearMiss-' + str(version) + ' ...'
undersampler = NearMiss(verbose=verbose, version=version)
X_undersampled, y_undersampled = undersampler.fit_transform(X, y)
return X_undersampled, y_undersampled
def nearmiss_undersampling(X,y,version): """ Perform NearMiss undersampling Keyword arguments: X -- The feature vectors y -- The target classes """ if verbose: print '\nUndersampling with NearMiss-'+str(version)+' ...' undersampler=NearMiss(verbose=verbose,version=version) X_undersampled,y_undersampled = undersampler.fit_transform(X,y) return X_undersampled,y_undersampled
def test_nm1_fit():
"""Test the fitting method"""
# Define the parameter for the under-sampling
ratio = 'auto'
# Create the object
nm1 = NearMiss(ratio=ratio, random_state=RND_SEED,
version=VERSION_NEARMISS)
# Fit the data
nm1.fit(X, Y)
# Check if the data information have been computed
assert_equal(nm1.min_c_, 0)
assert_equal(nm1.maj_c_, 1)
assert_equal(nm1.stats_c_[0], 500)
assert_equal(nm1.stats_c_[1], 4500)
def test_nm_fit_invalid_ratio():
"""Test either if an error is raised when the balancing ratio to fit is
smaller than the one of the data"""
# Create the object
ratio = 1. / 10000.
nm = NearMiss(ratio=ratio, random_state=RND_SEED)
# Fit the data
assert_raises(RuntimeError, nm.fit, X, Y)
def test_nm1_fit_transform_half():
"""Test fit and transform routines with .5 ratio"""
# Define the parameter for the under-sampling
ratio = .5
# Create the object
nm1 = NearMiss(ratio=ratio, random_state=RND_SEED,
version=VERSION_NEARMISS)
# Fit and transform
X_resampled, y_resampled = nm1.fit_transform(X, Y)
currdir = os.path.dirname(os.path.abspath(__file__))
X_gt = np.load(os.path.join(currdir, 'data', 'nm1_x_05.npy'))
y_gt = np.load(os.path.join(currdir, 'data', 'nm1_y_05.npy'))
assert_array_equal(X_resampled, X_gt)
assert_array_equal(y_resampled, y_gt)
def test_nm1_fit_transform_auto_indices():
"""Test fit and transform routines with auto ratio and indices support"""
# Define the parameter for the under-sampling
ratio = 'auto'
# Create the object
nm1 = NearMiss(ratio=ratio, random_state=RND_SEED,
version=VERSION_NEARMISS, return_indices=True)
# Fit and transform
X_resampled, y_resampled, idx_under = nm1.fit_transform(X, Y)
currdir = os.path.dirname(os.path.abspath(__file__))
X_gt = np.load(os.path.join(currdir, 'data', 'nm1_x.npy'))
y_gt = np.load(os.path.join(currdir, 'data', 'nm1_y.npy'))
idx_gt = np.load(os.path.join(currdir, 'data', 'nm1_idx.npy'))
assert_array_equal(X_resampled, X_gt)
assert_array_equal(y_resampled, y_gt)
assert_array_equal(idx_under, idx_gt)
def test_nm1_transform_wt_fit():
"""Test either if an error is raised when transform is called before
fitting"""
# Define the parameter for the under-sampling
ratio = 'auto'
# Create the object
nm1 = NearMiss(ratio=ratio, random_state=RND_SEED,
version=VERSION_NEARMISS)
assert_raises(RuntimeError, nm1.transform, X, Y)
def test_nearmiss_fit_single_class():
"""Test either if an error when there is a single class"""
# Define the parameter for the under-sampling
ratio = 'auto'
# Create the object
nm1 = NearMiss(ratio=ratio, random_state=RND_SEED,
version=VERSION_NEARMISS)
# Resample the data
# Create a wrong y
y_single_class = np.zeros((X.shape[0], ))
assert_raises(RuntimeError, nm1.fit, X, y_single_class)
def test_rest(x, y,c=0,ratio='auto'):
c=c
if(c==0):
print('Random under-sampling')
US = UnderSampler(indices_support=indices_support, verbose=verbose,ratio=ratio)
x, y, idx_tmp = US.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
elif(c==1):
print('Tomek links')
TL = TomekLinks(verbose=verbose,ratio=ratio)
x, y = TL.fit_transform(x, y)
elif(c==2):
print('Clustering centroids')
CC = ClusterCentroids(verbose=verbose,ratio=ratio)
x, y = CC.fit_transform(x, y)
elif(c==3):
print('NearMiss-1')
NM1 = NearMiss(version=1, indices_support=indices_support, verbose=verbose,ratio=ratio)
x, y, idx_tmp = NM1.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
elif(c==4):
print('NearMiss-2')
NM2 = NearMiss(version=2, indices_support=indices_support, verbose=verbose,ratio=ratio)
x, y, idx_tmp = NM2.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
elif(c==5):
print('NearMiss-3')
NM3 = NearMiss(version=3, indices_support=indices_support, verbose=verbose,ratio=ratio)
x, y, idx_tmp = NM3.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
elif(c==6):
print('Neighboorhood Cleaning Rule')
NCR = NeighbourhoodCleaningRule(indices_support=indices_support, verbose=verbose,ratio=ratio)
x, y, idx_tmp = NCR.fit_transform(x, y)
print ('Indices selected')
print(idx_tmp)
elif(c==7):
print('Random over-sampling')
OS = OverSampler(verbose=verbose,ratio=ratio)
x, y = OS.fit_transform(x, y)
elif(c==8):
print('SMOTE Tomek links')
STK = SMOTETomek(verbose=verbose,ratio=ratio)
x, y = STK.fit_transform(x, y)
elif(c==9):
print('SMOTE ENN')
SENN = SMOTEENN(verbose=verbose,ratio=ratio)
x, y = SENN.fit_transform(x, y)
else:
print('EasyEnsemble')
EE = EasyEnsemble(verbose=verbose,ratio=ratio)
x, y = EE.fit_transform(x, y)
return x, y
def test_nearmiss_init():
"""Test the initialisation of the object"""
# Define a ratio
ratio = 1.
verbose = True
nm1 = NearMiss(ratio=ratio, random_state=RND_SEED, verbose=verbose,
version=VERSION_NEARMISS)
assert_equal(nm1.version, VERSION_NEARMISS)
assert_equal(nm1.size_ngh, 3)
assert_equal(nm1.ratio_, ratio)
assert_equal(nm1.rs_, RND_SEED)
assert_equal(nm1.verbose, verbose)
assert_equal(nm1.min_c_, None)
assert_equal(nm1.maj_c_, None)
assert_equal(nm1.stats_c_, {})
def __get_sample_transformed_examples(sample_type, train_x, train_y, ratio):
sampler = None
verbose = True
if sample_type == SMOTE_REG:
sampler = SMOTE(kind='regular', verbose=verbose, ratio=ratio, k=15)
elif sample_type == SMOTE_SVM:
# TODO: Make this configurable?
svm_args = {'class_weight': 'balanced'}
sampler = SMOTE(kind='svm',
ratio=ratio,
verbose=verbose,
k=15,
**svm_args)
elif sample_type == SMOTE_BORDERLINE_1:
sampler = SMOTE(kind='borderline1', ratio=ratio, verbose=verbose)
elif sample_type == SMOTE_BORDERLINE_2:
sampler = SMOTE(kind='borderline2', ratio=ratio, verbose=verbose)
elif sample_type == SMOTE_ENN:
sampler = SMOTEENN(ratio=ratio, verbose=verbose, k=15)
elif sample_type == SMOTE_TOMEK:
sampler = SMOTETomek(ratio=ratio, verbose=verbose, k=15)
elif sample_type == UNDERSAMPLER:
sampler = UnderSampler(ratio=ratio,
verbose=verbose,
replacement=False,
random_state=17)
elif sample_type == ADASYN_SAMPLER:
sampler = ADASYN(k=15, imb_threshold=0.6, ratio=ratio)
elif sample_type == TOMEK_LINKS:
sampler = TomekLinks()
elif sample_type == CLUSTER_CENTROIDS:
sampler = ClusterCentroids(ratio=ratio)
elif sample_type == NEARMISS:
sampler = NearMiss(ratio=ratio)
else:
print "Unrecoqnized sample technique: " + sample_type
print "Returning original data"
return train_x, train_y
return sampler.fit_transform(train_x, train_y)
from unbalanced_dataset.under_sampling import NearMiss
# Generate the dataset
X, y = make_classification(n_classes=2, class_sep=2, weights=[0.1, 0.9],
n_informative=3, n_redundant=1, flip_y=0,
n_features=20, n_clusters_per_class=1,
n_samples=5000, random_state=10)
# Instanciate a PCA object for the sake of easy visualisation
pca = PCA(n_components=2)
# Fit and transform x to visualise inside a 2D feature space
X_vis = pca.fit_transform(X)
# Apply the random under-sampling
nm3 = NearMiss(version=1)
X_resampled, y_resampled = nm3.fit_transform(X, y)
X_res_vis = pca.transform(X_resampled)
# Two subplots, unpack the axes array immediately
f, (ax1, ax2) = plt.subplots(1, 2)
ax1.scatter(X_vis[y == 0, 0], X_vis[y == 0, 1], label="Class #0", alpha=0.5,
edgecolor=almost_black, facecolor=palette[0], linewidth=0.15)
ax1.scatter(X_vis[y == 1, 0], X_vis[y == 1, 1], label="Class #1", alpha=0.5,
edgecolor=almost_black, facecolor=palette[2], linewidth=0.15)
ax1.set_title('Original set')
ax2.scatter(X_res_vis[y_resampled == 0, 0], X_res_vis[y_resampled == 0, 1],
label="Class #0", alpha=.5, edgecolor=almost_black,
facecolor=palette[0], linewidth=0.15)
weights=[0.1, 0.9],
n_informative=3,
n_redundant=1,
flip_y=0,
n_features=20,
n_clusters_per_class=1,
n_samples=5000,
random_state=10)
# Instanciate a PCA object for the sake of easy visualisation
pca = PCA(n_components=2)
# Fit and transform x to visualise inside a 2D feature space
X_vis = pca.fit_transform(X)
# Apply the random under-sampling
nm3 = NearMiss(version=1)
X_resampled, y_resampled = nm3.fit_transform(X, y)
X_res_vis = pca.transform(X_resampled)
# Two subplots, unpack the axes array immediately
f, (ax1, ax2) = plt.subplots(1, 2)
ax1.scatter(X_vis[y == 0, 0],
X_vis[y == 0, 1],
label="Class #0",
alpha=0.5,
edgecolor=almost_black,
facecolor=palette[0],
linewidth=0.15)
ax1.scatter(X_vis[y == 1, 0],
X_vis[y == 1, 1],
from unbalanced_dataset.under_sampling import NearMiss
# Generate the dataset
X, y = make_classification(n_classes=2, class_sep=2, weights=[0.1, 0.9],
n_informative=3, n_redundant=1, flip_y=0,
n_features=20, n_clusters_per_class=1,
n_samples=5000, random_state=10)
# Instanciate a PCA object for the sake of easy visualisation
pca = PCA(n_components=2)
# Fit and transform x to visualise inside a 2D feature space
X_vis = pca.fit_transform(X)
# Apply the random under-sampling
nm2 = NearMiss(version=2)
X_resampled, y_resampled = nm2.fit_transform(X, y)
X_res_vis = pca.transform(X_resampled)
# Two subplots, unpack the axes array immediately
f, (ax1, ax2) = plt.subplots(1, 2)
ax1.scatter(X_vis[y == 0, 0], X_vis[y == 0, 1], label="Class #0", alpha=0.5,
edgecolor=almost_black, facecolor=palette[0], linewidth=0.15)
ax1.scatter(X_vis[y == 1, 0], X_vis[y == 1, 1], label="Class #1", alpha=0.5,
edgecolor=almost_black, facecolor=palette[2], linewidth=0.15)
ax1.set_title('Original set')
ax2.scatter(X_res_vis[y_resampled == 0, 0], X_res_vis[y_resampled == 0, 1],
label="Class #0", alpha=.5, edgecolor=almost_black,
facecolor=palette[0], linewidth=0.15)
