def feat_importance(X, cont, clf=None, n_estimators=1000, n_splits=10, max_samples=1.,
num_threads=24, pct_embargo=0., scoring='accuracy',
method='SFI', min_w_leaf=0., **kwargs):
n_jobs = (-1 if num_threads > 1 else 1)
# Build classifiers
if clf is None:
base_clf = DecisionTreeClassifier(criterion='entropy', max_features=1,
class_weight='balanced',
min_weight_fraction_leaf=min_w_leaf)
clf = BaggingClassifier(base_estimator=base_clf, n_estimators=n_estimators,
max_features=1., max_samples=max_samples,
oob_score=True, n_jobs=n_jobs)
fit_clf = clf.fit(X, cont['bin'], sample_weight=cont['w'].values)
if hasattr(fit_clf, 'oob_score_'):
oob = fit_clf.oob_score_
else:
oob = None
if method == 'MDI':
imp = feat_imp_MDI(fit_clf, feat_names=X.columns)
oos = cv_score(clf, X=X, y=cont['bin'], n_splits=n_splits,
sample_weight=cont['w'], t1=cont['t1'],
pct_embargo=pct_embargo, scoring=scoring).mean()
elif method == 'MDA':
imp, oos = feat_imp_MDA(clf, X=X, y=cont['bin'], n_splits=n_splits,
sample_weight=cont['w'], t1=cont['t1'],
pct_embargo=pct_embargo, scoring=scoring)
elif method == 'SFI':
cv_gen = PurgedKFold(n_splits=n_splits, t1=cont['t1'], pct_embargo=pct_embargo)
oos = cv_score(clf, X=X, y=cont['bin'], sample_weight=cont['w'],
scoring=scoring, cv_gen=cv_gen)
clf.n_jobs = 1
imp = mp_pandas_obj(aux_feat_imp_SFI, ('feat_names', X.columns),
num_threads, clf=clf, X=X, cont=cont,
scoring=scoring, cv_gen=cv_gen)
return imp, oob, oos
def get_sample_w(t1, num_co_events, close, num_threads=1):
"""
Snippet 4.10 (page 69) Determination Of Sample Weight By Absolute Return Attribution
:param t1:
:param num_co_events:
:param close:
:param num_threads:
:return:
"""
wght = mp_pandas_obj(mp_sample_w, ('molecule', t1.index), num_threads=num_threads,
t1=t1, num_co_events=num_co_events, close=close)
return wght
def get_sample_tw(t1, num_co_events, num_threads=1):
"""
Calculate sampling weight with considering some attributes
:param timestamps:
:param t1:
:param num_co_events:
:param num_threads:
:return:
"""
weight = mp_pandas_obj(mp_sample_tw, ('molecule', t1.index), num_threads=num_threads,
t1=t1, num_co_events=num_co_events)
return weight
def get_train_times(train_times, test_times, num_threads=1):
"""Sample train points without overlapping with test period
Params
------
train_times: pd.Series
Trainig points with index for initial and values for end time
test_times: pd.Series
Testing points with index for initial and values for end time
num_threads: int, default 1
The number of thrads for multiprocessing
Returns
-------
pd.Series
"""
return mp_pandas_obj(
mp_train_times, ('molecule', train_times.index),
num_threads,
train_times=train_times,
test_times=test_times)
def avg_active_signals(signals, num_threads=1, timestamps=None):
"""Average active signals
Paramters
---------
signals: pd.Series
num_threads: 1
timestamps: list, optional
Timestamps used for output. When there is not active signal,
value will be zero on that point. If not specified, use signals.index
Return
------
pd.Series
"""
if timestamps is None:
timestamps = set(signals['t1'].dropna().values)
timestamps = list(timestamps.union(set(signals.index.values)))
timestamps.sort()
out = mp_pandas_obj(mp_avg_active_signals, ('molecule', timestamps),
num_threads,
signals=signals)
return out