def calibration():
steps = []
for n_estimators, k_folds in product(range(50,300,100), [2,5]):
d = data.ClassificationData(target=True, n_samples=1000, n_features=100)
est = step.Construct('sklearn.ensemble.RandomForestClassifier',
n_estimators=n_estimators, name='estimator')
fit = model.Fit(inputs=[est, d], return_estimator=True, target=True, name='uncalibrated')
predict = model.Predict(inputs=[fit,d], target=True, name='y')
cal = step.Construct('sklearn.calibration.CalibratedClassifierCV', cv=k_folds,
inputs=[predict], inputs_mapping={'y':None}, name='calibrator')
cal_est = model.FitPredict(inputs=[cal, d], target=True, name='calibrated')
metrics = model.PrintMetrics([
{'metric':'baseline'},
{'metric':'precision', 'k':100},
{'metric':'precision', 'k':200},
{'metric':'precision', 'k':300},
], inputs=[cal_est])
steps.append(metrics)
return steps
def calibration():
steps = []
for n_estimators, k_folds in product(range(50, 300, 100), [2, 5]):
d = data.ClassificationData(n_samples=1000, n_features=100)
d.target = True
est = step.Call(ensemble,
'RandomForestClassifier',
n_estimators=n_estimators)
fit = model.Fit(inputs=[est, d], return_estimator=True)
fit.target = True
predict = model.Predict(inputs=[fit, d])
predict.target = True
cal = step.Call('sklearn.calibration.CalibratedClassifierCV',
cv=k_folds,
inputs=[MapResults([predict], {'y': None})])
cal_est = model.FitPredict(inputs=[cal, d])
cal_est.target = True
steps.append(cal_est)
return steps
def n_estimators_search():
d = data.ClassificationData(n_samples=1000, n_features=100)
d.target = True
predict = []
for n_estimators in range(1, 4):
e = step.Construct(_class='sklearn.ensemble.RandomForestClassifier',
n_estimators=n_estimators)
f = model.Fit(inputs=[e, d], return_estimator=True, return_feature_importances=True)
p = model.Predict(inputs=[f, d])
p.target = True
predict.append(p)
return predict
def prediction():
# generate the data including a training and test split
d = data.ClassificationData(n_samples=1000, n_features=100)
d.target = True
# construct a random forest estimator
e = step.Construct(_class='sklearn.ensemble.RandomForestClassifier', n_estimators=1)
e.target = False
# fit the estimator
f = model.Fit(inputs=[e, d], return_estimator=True, return_feature_importances=True)
# make predictions
p = model.Predict(inputs=[f, d])
p.target = True
return p
def models(estimators, cv_search, transform_search):
"""
Grid search prediction workflows. Used by bll6_models, test_models, and product_models.
Args:
estimators: collection of steps, each of which constructs an estimator
cv_search: dictionary of arguments to LeadCrossValidate to search over
transform_search: dictionary of arguments to LeadTransform to search over
Returns: a list drain.model.Predict steps constructed by taking the product of
the estimators with the the result of drain.util.dict_product on each of
cv_search and transform_search.
Each Predict step contains the following in its inputs graph:
- lead.model.cv.LeadCrossValidate
- lead.model.transform.LeadTransform
- drain.model.Fit
"""
steps = []
for cv_args, transform_args, estimator in product(
dict_product(cv_search), dict_product(transform_search),
estimators):
cv = lead.model.cv.LeadCrossValidate(**cv_args)
cv.name = 'cv'
X_train = Call('__getitem__',
inputs=[
MapResults(
[cv], {
'X': 'obj',
'train': 'key',
'test': None,
'aux': None
})
])
mean = Call('mean', inputs=[X_train])
mean.name = 'mean'
X_impute = Construct(data.impute,
inputs=[
MapResults([cv], {
'aux': None,
'test': None,
'train': None
}),
MapResults([mean], 'value')
])
cv_imputed = MapResults([X_impute, cv], ['X', {'X': None}])
cv_imputed.target = True
transform = lead.model.transform.LeadTransform(inputs=[cv_imputed],
**transform_args)
transform.name = 'transform'
fit = model.Fit(inputs=[estimator, transform], return_estimator=True)
fit.name = 'fit'
y = model.Predict(inputs=[fit, transform],
return_feature_importances=True)
y.name = 'predict'
y.target = True
steps.append(y)
return steps