def _parallel_fit(window_size, ):
rng = check_random_state(window_size)
all_words = [dict() for x in range(len(X))]
relevant_features_count = 0
# for window_size in self.window_sizes:
transformer = SFA(
word_length=rng.choice(self.word_lengths),
alphabet_size=self.alphabet_size,
window_size=window_size,
norm=rng.choice(self.norm_options),
anova=self.anova,
# levels=rng.choice([1, 2, 3]),
binning_method=self.binning_strategy,
bigrams=self.bigrams,
remove_repeat_words=False,
lower_bounding=False,
save_words=False,
)
sfa_words = transformer.fit_transform(X, y)
# self.SFA_transformers.append(transformer)
bag = sfa_words[0]
apply_chi_squared = self.p_threshold < 1
# chi-squared test to keep only relevant features
if apply_chi_squared:
vectorizer = DictVectorizer(sparse=True,
dtype=np.int32,
sort=False)
bag_vec = vectorizer.fit_transform(bag)
chi2_statistics, p = chi2(bag_vec, y)
relevant_features_idx = np.where(p <= self.p_threshold)[0]
relevant_features = set(
np.array(vectorizer.feature_names_)[relevant_features_idx])
relevant_features_count += len(relevant_features_idx)
# merging bag-of-patterns of different window_sizes
# to single bag-of-patterns with prefix indicating
# the used window-length
for j in range(len(bag)):
for (key, value) in bag[j].items():
# chi-squared test
if (not apply_chi_squared) or (key
in relevant_features):
# append the prefixes to the words to
# distinguish between window-sizes
word = WEASEL._shift_left(key, self.highest_bit,
window_size)
all_words[j][word] = value
return all_words, transformer, relevant_features_count
class IndividualBOSS(BaseClassifier):
"""Single bag of Symbolic Fourier Approximation Symbols (IndividualBOSS).
Bag of SFA Symbols Ensemble: implementation of a single BOSS Schaffer, the base
classifier for the boss ensemble.
Implementation of single BOSS model from Schäfer (2015). [1]_
This is the underlying classifier for each classifier in the BOSS ensemble.
Overview: input "n" series of length "m" and IndividualBoss performs a SFA
transform to form a sparse dictionary of discretised words. The resulting
dictionary is used with the BOSS distance function in a 1-nearest neighbor.
Fit involves finding "n" histograms.
Predict uses 1 nearest neighbor with a bespoke BOSS distance function.
Parameters
----------
window_size : int
Size of the window to use in BOSS algorithm.
word_length : int
Length of word to use to use in BOSS algorithm.
norm : bool, default = False
Whether to normalize words by dropping the first Fourier coefficient.
alphabet_size : default = 4
Number of possible letters (values) for each word.
save_words : bool, default = True
Whether to keep NumPy array of words in SFA transformation even after
the dictionary of words is returned. If True, the array is saved, which
can shorten the time to calculate dictionaries using a shorter
`word_length` (since the last "n" letters can be removed).
typed_dict : bool, default=True
Use a numba TypedDict to store word counts. May increase memory usage, but will
be faster for larger datasets.
n_jobs : int, default=1
The number of jobs to run in parallel for both `fit` and `predict`.
``-1`` means using all processors.
random_state : int or None, default=None
Seed for random, integer.
Attributes
----------
n_classes_ : int
Number of classes. Extracted from the data.
classes_ : list
The classes labels.
See Also
--------
BOSSEnsemble, ContractableBOSS
Notes
-----
For the Java version, see
`TSML <https://github.com/uea-machine-learning/tsml/blob/master/src/main/java/
tsml/classifiers/dictionary_based/IndividualBOSS.java>`_.
References
----------
.. [1] Patrick Schäfer, "The BOSS is concerned with time series classification
in the presence of noise", Data Mining and Knowledge Discovery, 29(6): 2015
https://link.springer.com/article/10.1007/s10618-014-0377-7
Examples
--------
>>> from sktime.classification.dictionary_based import IndividualBOSS
>>> from sktime.datasets import load_unit_test
>>> X_train, y_train = load_unit_test(split="train", return_X_y=True)
>>> X_test, y_test = load_unit_test(split="test", return_X_y=True)
>>> clf = IndividualBOSS()
>>> clf.fit(X_train, y_train)
IndividualBOSS(...)
>>> y_pred = clf.predict(X_test)
"""
_tags = {
"capability:multithreading": True,
}
def __init__(
self,
window_size=10,
word_length=8,
norm=False,
alphabet_size=4,
save_words=False,
typed_dict=True,
n_jobs=1,
random_state=None,
):
self.window_size = window_size
self.word_length = word_length
self.norm = norm
self.alphabet_size = alphabet_size
self.save_words = save_words
self.typed_dict = typed_dict
self.n_jobs = n_jobs
self.random_state = random_state
self._transformer = None
self._transformed_data = []
self._class_vals = []
self._accuracy = 0
self._subsample = []
self._train_predictions = []
super(IndividualBOSS, self).__init__()
def __getstate__(self):
"""Return state as dictionary for pickling, required for typed Dict objects."""
state = self.__dict__.copy()
if self.typed_dict:
nl = [None] * len(self._transformed_data)
for i, ndict in enumerate(state["_transformed_data"]):
pdict = dict()
for key, val in ndict.items():
pdict[key] = val
nl[i] = pdict
state["_transformed_data"] = nl
return state
def __setstate__(self, state):
"""Set current state using input pickling, required for typed Dict objects."""
self.__dict__.update(state)
if self.typed_dict:
nl = [None] * len(self._transformed_data)
for i, pdict in enumerate(self._transformed_data):
ndict = Dict.empty(key_type=types.int64, value_type=types.uint32)
for key, val in pdict.items():
ndict[key] = val
nl[i] = ndict
self._transformed_data = nl
def _fit(self, X, y):
"""Fit a single boss classifier on n_instances cases (X,y).
Parameters
----------
X : 3D np.array of shape = [n_instances, n_dimensions, series_length]
The training data.
y : array-like, shape = [n_instances]
The class labels.
Returns
-------
self :
Reference to self.
Notes
-----
Changes state by creating a fitted model that updates attributes
ending in "_" and sets is_fitted flag to True.
"""
self._transformer = SFA(
word_length=self.word_length,
alphabet_size=self.alphabet_size,
window_size=self.window_size,
norm=self.norm,
remove_repeat_words=True,
bigrams=False,
save_words=self.save_words,
typed_dict=self.typed_dict,
n_jobs=self._threads_to_use,
)
sfa = self._transformer.fit_transform(X)
self._transformed_data = sfa[0]
self._class_vals = y
return self
def _predict(self, X):
"""Predict class values of all instances in X.
Parameters
----------
X : 3D np.array of shape = [n_instances, n_dimensions, series_length]
The data to make predictions for.
Returns
-------
y : array-like, shape = [n_instances]
Predicted class labels.
"""
test_bags = self._transformer.transform(X)
test_bags = test_bags[0]
classes = Parallel(n_jobs=self._threads_to_use)(
delayed(self._test_nn)(
test_bag,
)
for test_bag in test_bags
)
return np.array(classes)
def _test_nn(self, test_bag):
rng = check_random_state(self.random_state)
best_dist = sys.float_info.max
nn = None
for n, bag in enumerate(self._transformed_data):
dist = boss_distance(test_bag, bag, best_dist)
if dist < best_dist or (dist == best_dist and rng.random() < 0.5):
best_dist = dist
nn = self._class_vals[n]
return nn
def _train_predict(self, train_num):
test_bag = self._transformed_data[train_num]
best_dist = sys.float_info.max
nn = None
for n, bag in enumerate(self._transformed_data):
if n == train_num:
continue
dist = boss_distance(test_bag, bag, best_dist)
if dist < best_dist:
best_dist = dist
nn = self._class_vals[n]
return nn
def _shorten_bags(self, word_len):
new_boss = IndividualBOSS(
self.window_size,
word_len,
self.norm,
self.alphabet_size,
save_words=self.save_words,
typed_dict=self.typed_dict,
random_state=self.random_state,
n_jobs=self.n_jobs,
)
new_boss._transformer = self._transformer
sfa = self._transformer._shorten_bags(word_len)
new_boss._transformed_data = sfa[0]
new_boss._class_vals = self._class_vals
new_boss.n_classes_ = self.n_classes_
new_boss.classes_ = self.classes_
new_boss._class_dictionary = self._class_dictionary
new_boss._threads_to_use = self._threads_to_use
new_boss._is_fitted = True
return new_boss
def _clean(self):
self._transformer.words = None
self._transformer.save_words = False
def _set_word_len(self, word_len):
self.word_length = word_len
self._transformer.word_length = word_len
def _fit(self, X, y):
"""Build a WEASEL+MUSE classifiers from the training set (X, y).
Parameters
----------
X : nested pandas DataFrame of shape [n_instances, 1]
Nested dataframe with univariate time-series in cells.
y : array-like, shape = [n_instances] The class labels.
Returns
-------
self : object
"""
y = np.asarray(y)
self.classes_ = class_distribution(np.asarray(y).reshape(-1, 1))[0][0]
# add first order differences in each dimension to TS
if self.use_first_order_differences:
X = self._add_first_order_differences(X)
# Window length parameter space dependent on series length
self.col_names = X.columns
rng = check_random_state(self.random_state)
self.n_dims = len(self.col_names)
self.highest_dim_bit = (math.ceil(math.log2(self.n_dims))) + 1
self.highest_bits = np.zeros(self.n_dims)
self.SFA_transformers = [[] for _ in range(self.n_dims)]
# the words of all dimensions and all time series
all_words = [dict() for _ in range(X.shape[0])]
# On each dimension, perform SFA
for ind, column in enumerate(self.col_names):
X_dim = X[[column]]
X_dim = from_nested_to_3d_numpy(X_dim)
series_length = X_dim.shape[
-1] # TODO compute minimum over all ts ?
# increment window size in steps of 'win_inc'
win_inc = self._compute_window_inc(series_length)
self.max_window = int(min(series_length, self.max_window))
if self.min_window > self.max_window:
raise ValueError(
f"Error in MUSE, min_window ="
f"{self.min_window} is bigger"
f" than max_window ={self.max_window},"
f" series length is {self.series_length}"
f" try set min_window to be smaller than series length in "
f"the constructor, but the classifier may not work at "
f"all with very short series")
self.window_sizes.append(
list(range(self.min_window, self.max_window, win_inc)))
self.highest_bits[ind] = math.ceil(math.log2(self.max_window)) + 1
for window_size in self.window_sizes[ind]:
transformer = SFA(
word_length=rng.choice(self.word_lengths),
alphabet_size=self.alphabet_size,
window_size=window_size,
norm=rng.choice(self.norm_options),
anova=self.anova,
binning_method=rng.choice(self.binning_strategies),
bigrams=self.bigrams,
remove_repeat_words=False,
lower_bounding=False,
save_words=False,
)
sfa_words = transformer.fit_transform(X_dim, y)
self.SFA_transformers[ind].append(transformer)
bag = sfa_words[0]
# chi-squared test to keep only relevant features
relevant_features = {}
apply_chi_squared = self.p_threshold < 1
if apply_chi_squared:
vectorizer = DictVectorizer(sparse=True,
dtype=np.int32,
sort=False)
bag_vec = vectorizer.fit_transform(bag)
chi2_statistics, p = chi2(bag_vec, y)
relevant_features_idx = np.where(p <= self.p_threshold)[0]
relevant_features = set(
np.array(
vectorizer.feature_names_)[relevant_features_idx])
# merging bag-of-patterns of different window_sizes
# to single bag-of-patterns with prefix indicating
# the used window-length
highest = np.int32(self.highest_bits[ind])
for j in range(len(bag)):
for (key, value) in bag[j].items():
# chi-squared test
if (not apply_chi_squared) or (key
in relevant_features):
# append the prefices to the words to
# distinguish between window-sizes
word = MUSE._shift_left(key, highest, ind,
self.highest_dim_bit,
window_size)
all_words[j][word] = value
self.clf = make_pipeline(
DictVectorizer(sparse=True, sort=False),
# StandardScaler(with_mean=True, copy=False),
LogisticRegression(
max_iter=5000,
solver="liblinear",
dual=True,
# class_weight="balanced",
penalty="l2",
random_state=self.random_state,
),
)
self.clf.fit(all_words, y)
return self
class IndividualBOSS(BaseClassifier):
"""Single Bag of SFA Symbols (BOSS) classifier
Bag of SFA Symbols Ensemble: implementation of BOSS from Schaffer :
@article
"""
def __init__(
self,
window_size=10,
word_length=8,
norm=False,
alphabet_size=4,
save_words=True,
random_state=None,
):
self.window_size = window_size
self.word_length = word_length
self.norm = norm
self.alphabet_size = alphabet_size
self.save_words = save_words
self.random_state = random_state
self.transformer = SFA(
word_length=word_length,
alphabet_size=alphabet_size,
window_size=window_size,
norm=norm,
remove_repeat_words=True,
bigrams=False,
save_words=save_words,
)
self.transformed_data = []
self.accuracy = 0
self.class_vals = []
self.num_classes = 0
self.classes_ = []
self.class_dictionary = {}
super(IndividualBOSS, self).__init__()
def fit(self, X, y):
X, y = check_X_y(X, y, enforce_univariate=True, coerce_to_numpy=True)
sfa = self.transformer.fit_transform(X)
self.transformed_data = sfa[0] # .iloc[:, 0]
self.class_vals = y
self.num_classes = np.unique(y).shape[0]
self.classes_ = class_distribution(np.asarray(y).reshape(-1, 1))[0][0]
for index, classVal in enumerate(self.classes_):
self.class_dictionary[classVal] = index
self._is_fitted = True
return self
def predict(self, X):
self.check_is_fitted()
X = check_X(X, enforce_univariate=True, coerce_to_numpy=True)
rng = check_random_state(self.random_state)
classes = []
test_bags = self.transformer.transform(X)
test_bags = test_bags[0] # .iloc[:, 0]
for test_bag in test_bags:
best_dist = sys.float_info.max
nn = None
for n, bag in enumerate(self.transformed_data):
dist = boss_distance(test_bag, bag, best_dist)
if dist < best_dist or (dist == best_dist and rng.random() < 0.5):
best_dist = dist
nn = self.class_vals[n]
classes.append(nn)
return np.array(classes)
def predict_proba(self, X):
preds = self.predict(X)
dists = np.zeros((X.shape[0], self.num_classes))
for i in range(0, X.shape[0]):
dists[i, self.class_dictionary.get(preds[i])] += 1
return dists
def _train_predict(self, train_num):
test_bag = self.transformed_data[train_num]
best_dist = sys.float_info.max
nn = None
for n, bag in enumerate(self.transformed_data):
if n == train_num:
continue
dist = boss_distance(test_bag, bag, best_dist)
if dist < best_dist:
best_dist = dist
nn = self.class_vals[n]
return nn
def _shorten_bags(self, word_len):
new_boss = IndividualBOSS(
self.window_size,
word_len,
self.norm,
self.alphabet_size,
save_words=self.save_words,
random_state=self.random_state,
)
new_boss.transformer = self.transformer
sfa = self.transformer._shorten_bags(word_len)
new_boss.transformed_data = sfa[0] # .iloc[:, 0]
new_boss.class_vals = self.class_vals
new_boss.num_classes = self.num_classes
new_boss.classes_ = self.classes_
new_boss.class_dictionary = self.class_dictionary
new_boss._is_fitted = True
return new_boss
def _clean(self):
self.transformer.words = None
self.transformer.save_words = False
def _set_word_len(self, word_len):
self.word_length = word_len
self.transformer.word_length = word_len
class IndividualBOSS(BaseClassifier):
"""Single Bag of SFA Symbols (BOSS) classifier.
Bag of SFA Symbols Ensemble: implementation of a single BOSS Schaffer, the base
classifier for the boss ensemble.
"""
def __init__(
self,
window_size=10,
word_length=8,
norm=False,
alphabet_size=4,
save_words=True,
n_jobs=1,
random_state=None,
):
self.window_size = window_size
self.word_length = word_length
self.norm = norm
self.alphabet_size = alphabet_size
self.save_words = save_words
self.n_jobs = n_jobs
self.random_state = random_state
self.transformer = SFA(
word_length=word_length,
alphabet_size=alphabet_size,
window_size=window_size,
norm=norm,
remove_repeat_words=True,
bigrams=False,
save_words=save_words,
n_jobs=n_jobs,
)
self.transformed_data = []
self.accuracy = 0
self.subsample = []
self.class_vals = []
self.num_classes = 0
self.classes_ = []
self.class_dictionary = {}
super(IndividualBOSS, self).__init__()
def fit(self, X, y):
"""Fit a single boss classifier on n_instances cases (X,y).
Parameters
----------
X : pd.DataFrame of shape [n_instances, 1]
Nested dataframe with univariate time-series in cells.
y : array-like, shape = [n_instances] The class labels.
Returns
-------
self : object
"""
X, y = check_X_y(X, y, enforce_univariate=True, coerce_to_numpy=True)
sfa = self.transformer.fit_transform(X)
self.transformed_data = sfa[0]
self.class_vals = y
self.num_classes = np.unique(y).shape[0]
self.classes_ = class_distribution(np.asarray(y).reshape(-1, 1))[0][0]
for index, classVal in enumerate(self.classes_):
self.class_dictionary[classVal] = index
self._is_fitted = True
return self
def predict(self, X):
"""Predict class values of all instances in X.
Parameters
----------
X : pd.DataFrame of shape [n, 1]
Returns
-------
array of shape [n, 1]
"""
self.check_is_fitted()
X = check_X(X, enforce_univariate=True, coerce_to_numpy=True)
test_bags = self.transformer.transform(X)
test_bags = test_bags[0]
classes = Parallel(n_jobs=self.n_jobs)(
delayed(self._test_nn)(test_bag, ) for test_bag in test_bags)
return np.array(classes)
def predict_proba(self, X):
"""Predict class probabilities for all instances in X.
Parameters
----------
X : pd.DataFrame of shape [n, 1]
Returns
-------
array of shape [n, self.n_classes]
"""
preds = self.predict(X)
dists = np.zeros((X.shape[0], self.num_classes))
for i in range(0, X.shape[0]):
dists[i, self.class_dictionary.get(preds[i])] += 1
return dists
def _test_nn(self, test_bag):
rng = check_random_state(self.random_state)
best_dist = sys.float_info.max
nn = None
for n, bag in enumerate(self.transformed_data):
dist = boss_distance(test_bag, bag, best_dist)
if dist < best_dist or (dist == best_dist and rng.random() < 0.5):
best_dist = dist
nn = self.class_vals[n]
return nn
def _train_predict(self, train_num):
test_bag = self.transformed_data[train_num]
best_dist = sys.float_info.max
nn = None
for n, bag in enumerate(self.transformed_data):
if n == train_num:
continue
dist = boss_distance(test_bag, bag, best_dist)
if dist < best_dist:
best_dist = dist
nn = self.class_vals[n]
return nn
def _shorten_bags(self, word_len):
new_boss = IndividualBOSS(
self.window_size,
word_len,
self.norm,
self.alphabet_size,
save_words=self.save_words,
random_state=self.random_state,
)
new_boss.transformer = self.transformer
sfa = self.transformer._shorten_bags(word_len)
new_boss.transformed_data = sfa[0]
new_boss.class_vals = self.class_vals
new_boss.num_classes = self.num_classes
new_boss.classes_ = self.classes_
new_boss.class_dictionary = self.class_dictionary
new_boss._is_fitted = True
return new_boss
def _clean(self):
self.transformer.words = None
self.transformer.save_words = False
def _set_word_len(self, word_len):
self.word_length = word_len
self.transformer.word_length = word_len
class IndividualTDE(BaseClassifier):
"""Single TDE classifier, based off the Bag of SFA Symbols (BOSS) model"""
def __init__(
self,
window_size=10,
word_length=8,
norm=False,
levels=1,
igb=False,
alphabet_size=4,
random_state=None,
):
self.window_size = window_size
self.word_length = word_length
self.norm = norm
self.levels = levels
self.igb = igb
self.alphabet_size = alphabet_size
self.random_state = random_state
binning_method = "information-gain" if igb else "equi-depth"
self.transformer = SFA(
word_length=word_length,
alphabet_size=alphabet_size,
window_size=window_size,
norm=norm,
levels=levels,
binning_method=binning_method,
bigrams=True,
remove_repeat_words=True,
save_words=False,
)
self.transformed_data = []
self.accuracy = 0
self.class_vals = []
self.num_classes = 0
self.classes_ = []
self.class_dictionary = {}
super(IndividualTDE, self).__init__()
def fit(self, X, y):
X, y = check_X_y(X, y, enforce_univariate=True, coerce_to_numpy=True)
sfa = self.transformer.fit_transform(X, y)
self.transformed_data = sfa[0] # .iloc[:, 0]
self.class_vals = y
self.num_classes = np.unique(y).shape[0]
self.classes_ = class_distribution(np.asarray(y).reshape(-1, 1))[0][0]
for index, classVal in enumerate(self.classes_):
self.class_dictionary[classVal] = index
self._is_fitted = True
return self
def predict(self, X):
self.check_is_fitted()
X = check_X(X, enforce_univariate=True, coerce_to_numpy=True)
rng = check_random_state(self.random_state)
classes = []
test_bags = self.transformer.transform(X)
test_bags = test_bags[0] # .iloc[:, 0]
for test_bag in test_bags:
best_sim = -1
nn = None
for n, bag in enumerate(self.transformed_data):
sim = histogram_intersection(test_bag, bag)
if sim > best_sim or (sim == best_sim and rng.random() < 0.5):
best_sim = sim
nn = self.class_vals[n]
classes.append(nn)
return np.array(classes)
def predict_proba(self, X):
preds = self.predict(X)
dists = np.zeros((X.shape[0], self.num_classes))
for i in range(0, X.shape[0]):
dists[i, self.class_dictionary.get(preds[i])] += 1
return dists
def _train_predict(self, train_num):
test_bag = self.transformed_data[train_num]
best_sim = -1
nn = None
for n, bag in enumerate(self.transformed_data):
if n == train_num:
continue
sim = histogram_intersection(test_bag, bag)
if sim > best_sim:
best_sim = sim
nn = self.class_vals[n]
return nn
class IndividualBOSS(BaseClassifier):
"""Single bag of Symbolic Fourier Approximation Symbols (IndividualBOSS).
Bag of SFA Symbols Ensemble: implementation of a single BOSS Schaffer, the base
classifier for the boss ensemble.
Implementation of single BOSS model from Schäfer (2015). [1]_
This is the underlying classifier for each classifier in the BOSS ensemble.
Overview: input "n" series of length "m" and IndividualBoss performs a SFA
transform to form a sparse dictionary of discretised words. The resulting
dictionary is used with the BOSS distance function in a 1-nearest neighbor.
Fit involves finding "n" histograms.
Predict uses 1 nearest neighbor with a bespoke BOSS distance function.
Parameters
----------
window_size : int
Size of the window to use in BOSS algorithm.
word_length : int
Length of word to use to use in BOSS algorithm.
norm : bool, default = False
Whether to normalize words by dropping the first Fourier coefficient.
alphabet_size : default = 4
Number of possible letters (values) for each word.
save_words : bool, default = True
Whether to keep NumPy array of words in SFA transformation even after
the dictionary of words is returned. If True, the array is saved, which
can shorten the time to calculate dictionaries using a shorter
`word_length` (since the last "n" letters can be removed).
n_jobs : int, default=1
The number of jobs to run in parallel for both `fit` and `predict`.
``-1`` means using all processors.
random_state : int or None, default=None
Seed for random, integer.
Attributes
----------
n_classes : int
Number of classes. Extracted from the data.
n_instances : int
Number of instances. Extracted from the data.
n_estimators : int
The final number of classifiers used. Will be <= `max_ensemble_size` if
`max_ensemble_size` has been specified.
series_length : int
Length of all series (assumed equal).
class_dictionary: dict
Dictionary of classes. Extracted from the data.
See Also
--------
BOSSEnsemble, ContractableBOSS
Notes
-----
For the Java version, see
`TSML <https://github.com/uea-machine-learning/tsml/blob/master/src/main/java/
tsml/classifiers/dictionary_based/BOSS.java>`_.
References
----------
.. [1] Patrick Schäfer, "The BOSS is concerned with time series classification
in the presence of noise", Data Mining and Knowledge Discovery, 29(6): 2015
https://link.springer.com/article/10.1007/s10618-014-0377-7
Examples
--------
>>> from sktime.classification.dictionary_based import IndividualBOSS
>>> from sktime.datasets import load_italy_power_demand
>>> X_train, y_train = load_italy_power_demand(split="train", return_X_y=True)
>>> X_test, y_test = load_italy_power_demand(split="test", return_X_y=True)
>>> clf = IndividualBOSS()
>>> clf.fit(X_train, y_train)
IndividualBOSS(...)
>>> y_pred = clf.predict(X_test)
"""
def __init__(
self,
window_size=10,
word_length=8,
norm=False,
alphabet_size=4,
save_words=True,
n_jobs=1,
random_state=None,
):
self.window_size = window_size
self.word_length = word_length
self.norm = norm
self.alphabet_size = alphabet_size
self.save_words = save_words
self.n_jobs = n_jobs
self.random_state = random_state
self.transformer = SFA(
word_length=word_length,
alphabet_size=alphabet_size,
window_size=window_size,
norm=norm,
remove_repeat_words=True,
bigrams=False,
save_words=save_words,
n_jobs=n_jobs,
)
self.transformed_data = []
self.accuracy = 0
self.subsample = []
self.class_vals = []
self.num_classes = 0
self.classes_ = []
self.class_dictionary = {}
super(IndividualBOSS, self).__init__()
def fit(self, X, y):
"""Fit a single boss classifier on n_instances cases (X,y).
Parameters
----------
X : pd.DataFrame of shape [n_instances, 1]
Nested dataframe with univariate time-series in cells.
y : array-like, shape = [n_instances] The class labels.
Returns
-------
self : object
"""
X, y = check_X_y(X, y, enforce_univariate=True, coerce_to_numpy=True)
sfa = self.transformer.fit_transform(X)
self.transformed_data = sfa[0]
self.class_vals = y
self.num_classes = np.unique(y).shape[0]
self.classes_ = class_distribution(np.asarray(y).reshape(-1, 1))[0][0]
for index, classVal in enumerate(self.classes_):
self.class_dictionary[classVal] = index
self._is_fitted = True
return self
def predict(self, X):
"""Predict class values of all instances in X.
Parameters
----------
X : pd.DataFrame of shape [n, 1]
Returns
-------
array of shape [n, 1]
"""
self.check_is_fitted()
X = check_X(X, enforce_univariate=True, coerce_to_numpy=True)
test_bags = self.transformer.transform(X)
test_bags = test_bags[0]
classes = Parallel(n_jobs=self.n_jobs)(
delayed(self._test_nn)(test_bag, ) for test_bag in test_bags)
return np.array(classes)
def predict_proba(self, X):
"""Predict class probabilities for all instances in X.
Parameters
----------
X : pd.DataFrame of shape [n, 1]
Returns
-------
dists : array of shape [n, self.n_classes]
"""
preds = self.predict(X)
dists = np.zeros((X.shape[0], self.num_classes))
for i in range(0, X.shape[0]):
dists[i, self.class_dictionary.get(preds[i])] += 1
return dists
def _test_nn(self, test_bag):
rng = check_random_state(self.random_state)
best_dist = sys.float_info.max
nn = None
for n, bag in enumerate(self.transformed_data):
dist = boss_distance(test_bag, bag, best_dist)
if dist < best_dist or (dist == best_dist and rng.random() < 0.5):
best_dist = dist
nn = self.class_vals[n]
return nn
def _train_predict(self, train_num):
test_bag = self.transformed_data[train_num]
best_dist = sys.float_info.max
nn = None
for n, bag in enumerate(self.transformed_data):
if n == train_num:
continue
dist = boss_distance(test_bag, bag, best_dist)
if dist < best_dist:
best_dist = dist
nn = self.class_vals[n]
return nn
def _shorten_bags(self, word_len):
new_boss = IndividualBOSS(
self.window_size,
word_len,
self.norm,
self.alphabet_size,
save_words=self.save_words,
random_state=self.random_state,
)
new_boss.transformer = self.transformer
sfa = self.transformer._shorten_bags(word_len)
new_boss.transformed_data = sfa[0]
new_boss.class_vals = self.class_vals
new_boss.num_classes = self.num_classes
new_boss.classes_ = self.classes_
new_boss.class_dictionary = self.class_dictionary
new_boss._is_fitted = True
return new_boss
def _clean(self):
self.transformer.words = None
self.transformer.save_words = False
def _set_word_len(self, word_len):
self.word_length = word_len
self.transformer.word_length = word_len