PREDICT.processing package¶
Submodules¶
PREDICT.processing.AdvancedSampler module¶
-
class
PREDICT.processing.AdvancedSampler.
AdvancedSampler
(param_distributions, n_iter, random_state=None, method='Halton')¶ Bases:
object
Generator on parameters sampled from given distributions using numerical sequences. Based on the sklearn ParameterSampler.
Non-deterministic iterable over random candidate combinations for hyper- parameter search. If all parameters are presented as a list, sampling without replacement is performed. If at least one parameter is given as a distribution, sampling with replacement is used. It is highly recommended to use continuous distributions for continuous parameters.
Note that before SciPy 0.16, the
scipy.stats.distributions
do not accept a custom RNG instance and always use the singleton RNG fromnumpy.random
. Hence settingrandom_state
will not guarantee a deterministic iteration wheneverscipy.stats
distributions are used to define the parameter search space. Deterministic behavior is however guaranteed from SciPy 0.16 onwards.Read more in the User Guide.
- param_distributions : dict
- Dictionary where the keys are parameters and values
are distributions from which a parameter is to be sampled.
Distributions either have to provide a
rvs
function to sample from them, or can be given as a list of values, where a uniform distribution is assumed. - n_iter : integer
- Number of parameter settings that are produced.
- random_state : int or RandomState
- Pseudo random number generator state used for random uniform sampling from lists of possible values instead of scipy.stats distributions.
- params : dict of string to any
- Yields dictionaries mapping each estimator parameter to as sampled value.
>>> from PREDICT.processing.HaltonSampler import HaltonSampler >>> from scipy.stats.distributions import expon >>> import numpy as np >>> np.random.seed(0) >>> param_grid = {'a':[1, 2], 'b': expon()} >>> param_list = list(HaltonSampler(param_grid, n_iter=4)) >>> rounded_list = [dict((k, round(v, 6)) for (k, v) in d.items()) ... for d in param_list] >>> rounded_list == [{'b': 0.89856, 'a': 1}, ... {'b': 0.923223, 'a': 1}, ... {'b': 1.878964, 'a': 2}, ... {'b': 1.038159, 'a': 2}] True
PREDICT.processing.ICC module¶
-
PREDICT.processing.ICC.
ICC
(M, ICCtype='inter')¶ - Input:
- M is matrix of observations. Rows: patients, columns: observers. type: ICC type, currently “inter” or “intra”.
-
PREDICT.processing.ICC.
ICC_anova
(Y, ICCtype='inter', more=False)¶ Adopted from Nipype with a slight alteration to distinguish inter and intra. the data Y are entered as a ‘table’ ie subjects are in rows and repeated measures in columns One Sample Repeated measure ANOVA Y = XB + E with X = [FaTor / Subjects]
PREDICT.processing.Imputer module¶
PREDICT.processing.SearchCV module¶
-
class
PREDICT.processing.SearchCV.
BaseSearchCV
(estimator, param_distributions={}, n_iter=10, scoring=None, fit_params=None, n_jobs=1, iid=True, refit=True, cv=None, verbose=0, pre_dispatch='2*n_jobs', random_state=None, error_score='raise', return_train_score=True, n_jobspercore=100, maxlen=100, fastr_plugin=None)¶ Bases:
abc.NewBase
Base class for hyper parameter search with cross-validation.
-
best_params_
¶
-
best_score_
¶
-
create_ensemble
(X_train, Y_train, verbose=None, initialize=True, scoring=None, method=50)¶ Create an (optimal) ensemble of a combination of hyperparameter settings and the associated groupsels, PCAs, estimators etc.
Based on Caruana et al. 2004, but a little different:
- Recreate the training/validation splits for a n-fold cross validation.
- For each fold:
- Start with an empty ensemble
- Create starting ensemble by adding N individually best performing models on the validation set. N is tuned on the validation set.
- Add model that improves ensemble performance on validation set the most, with replacement.
- Repeat (c) untill performance does not increase
The performance metric is the same as for the original hyperparameter search, i.e. probably the F1-score for classification and r2-score for regression. However, we recommend using the SAR score, as this is more universal.
Method: top50 or Caruana
-
decision_function
(*args, **kwargs)¶ Call decision_function on the estimator with the best found parameters.
Only available if
refit=True
and the underlying estimator supportsdecision_function
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
grid_scores_
¶
-
inverse_transform
(*args, **kwargs)¶ Call inverse_transform on the estimator with the best found params.
Only available if the underlying estimator implements
inverse_transform
andrefit=True
.- Xt : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
predict
(*args, **kwargs)¶ Call predict on the estimator with the best found parameters.
Only available if
refit=True
and the underlying estimator supportspredict
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
predict_log_proba
(*args, **kwargs)¶ Call predict_log_proba on the estimator with the best found parameters.
Only available if
refit=True
and the underlying estimator supportspredict_log_proba
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
predict_proba
(*args, **kwargs)¶ Call predict_proba on the estimator with the best found parameters.
Only available if
refit=True
and the underlying estimator supportspredict_proba
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
preprocess
(X)¶ Apply the available preprocssing methods to the features
-
process_fit
(n_splits, parameters_est, parameters_all, test_sample_counts, test_scores, train_scores, fit_time, score_time, cv_iter, base_estimator, X, y)¶ Process the outcomes of a SearchCV fit and find the best settings over all cross validations from all hyperparameters tested
-
refit_and_score
(X, y, parameters_all, parameters_est, train, test, verbose=None)¶ Refit the base estimator and attributes such as GroupSel
- X: array, mandatory
- Array containingfor each object (rows) the feature values (1st Column) and the associated feature label (2nd Column).
- y: list(?), mandatory
- List containing the labels of the objects.
- parameters_all: dictionary, mandatory
- Contains the settings used for the all preprocessing functions and the fitting. TODO: Create a default object and show the fields.
- parameters_est: dictionary, mandatory
- Contains the settings used for the base estimator
- train: list, mandatory
- Indices of the objects to be used as training set.
- test: list, mandatory
- Indices of the objects to be used as testing set.
-
score
(X, y=None)¶ Returns the score on the given data, if the estimator has been refit.
This uses the score defined by
scoring
where provided, and thebest_estimator_.score
method otherwise.- X : array-like, shape = [n_samples, n_features]
- Input data, where n_samples is the number of samples and n_features is the number of features.
- y : array-like, shape = [n_samples] or [n_samples, n_output], optional
- Target relative to X for classification or regression; None for unsupervised learning.
score : float
-
transform
(*args, **kwargs)¶ Call transform on the estimator with the best found parameters.
Only available if the underlying estimator supports
transform
andrefit=True
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
-
class
PREDICT.processing.SearchCV.
BaseSearchCVJoblib
(estimator, param_distributions={}, n_iter=10, scoring=None, fit_params=None, n_jobs=1, iid=True, refit=True, cv=None, verbose=0, pre_dispatch='2*n_jobs', random_state=None, error_score='raise', return_train_score=True, n_jobspercore=100, maxlen=100, fastr_plugin=None)¶ Bases:
PREDICT.processing.SearchCV.BaseSearchCV
Base class for hyper parameter search with cross-validation.
-
class
PREDICT.processing.SearchCV.
BaseSearchCVfastr
(estimator, param_distributions={}, n_iter=10, scoring=None, fit_params=None, n_jobs=1, iid=True, refit=True, cv=None, verbose=0, pre_dispatch='2*n_jobs', random_state=None, error_score='raise', return_train_score=True, n_jobspercore=100, maxlen=100, fastr_plugin=None)¶ Bases:
PREDICT.processing.SearchCV.BaseSearchCV
Base class for hyper parameter search with cross-validation.
-
class
PREDICT.processing.SearchCV.
Ensemble
(estimators)¶ Bases:
abc.NewBase
Ensemble of BaseSearchCV Estimators.
-
decision_function
(X)¶ Call decision_function on the estimator with the best found parameters.
Only available if
refit=True
and the underlying estimator supportsdecision_function
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
inverse_transform
(Xt)¶ Call inverse_transform on the estimator with the best found params.
Only available if the underlying estimator implements
inverse_transform
andrefit=True
.- Xt : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
predict
(X)¶ Call predict on the estimator with the best found parameters.
Only available if
refit=True
and the underlying estimator supportspredict
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
predict_log_proba
(X)¶ Call predict_log_proba on the estimator with the best found parameters.
Only available if
refit=True
and the underlying estimator supportspredict_log_proba
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
predict_proba
(X)¶ Call predict_proba on the estimator with the best found parameters.
Only available if
refit=True
and the underlying estimator supportspredict_proba
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
transform
(X)¶ Call transform on the estimator with the best found parameters.
Only available if the underlying estimator supports
transform
andrefit=True
.- X : indexable, length n_samples
- Must fulfill the input assumptions of the underlying estimator.
-
-
class
PREDICT.processing.SearchCV.
GridSearchCVJoblib
(estimator, param_grid, scoring=None, fit_params=None, n_jobs=1, iid=True, refit=True, cv=None, verbose=0, pre_dispatch='2*n_jobs', error_score='raise', return_train_score=True)¶ Bases:
PREDICT.processing.SearchCV.BaseSearchCVJoblib
Exhaustive search over specified parameter values for an estimator.
Important members are fit, predict.
GridSearchCV implements a “fit” and a “score” method. It also implements “predict”, “predict_proba”, “decision_function”, “transform” and “inverse_transform” if they are implemented in the estimator used.
The parameters of the estimator used to apply these methods are optimized by cross-validated grid-search over a parameter grid.
Read more in the User Guide.
- estimator : estimator object.
- This is assumed to implement the scikit-learn estimator interface.
Either estimator needs to provide a
score
function, orscoring
must be passed. - param_grid : dict or list of dictionaries
- Dictionary with parameters names (string) as keys and lists of parameter settings to try as values, or a list of such dictionaries, in which case the grids spanned by each dictionary in the list are explored. This enables searching over any sequence of parameter settings.
- scoring : string, callable or None, default=None
- A string (see model evaluation documentation) or
a scorer callable object / function with signature
scorer(estimator, X, y)
. IfNone
, thescore
method of the estimator is used. - fit_params : dict, optional
- Parameters to pass to the fit method.
- n_jobs : int, default=1
- Number of jobs to run in parallel.
- pre_dispatch : int, or string, optional
Controls the number of jobs that get dispatched during parallel execution. Reducing this number can be useful to avoid an explosion of memory consumption when more jobs get dispatched than CPUs can process. This parameter can be:
- None, in which case all the jobs are immediately created and spawned. Use this for lightweight and fast-running jobs, to avoid delays due to on-demand spawning of the jobs
- An int, giving the exact number of total jobs that are spawned
- A string, giving an expression as a function of n_jobs, as in ‘2*n_jobs’
- iid : boolean, default=True
- If True, the data is assumed to be identically distributed across the folds, and the loss minimized is the total loss per sample, and not the mean loss across the folds.
- cv : int, cross-validation generator or an iterable, optional
Determines the cross-validation splitting strategy. Possible inputs for cv are:
- None, to use the default 3-fold cross validation,
- integer, to specify the number of folds in a (Stratified)KFold,
- An object to be used as a cross-validation generator.
- An iterable yielding train, test splits.
For integer/None inputs, if the estimator is a classifier and
y
is either binary or multiclass,StratifiedKFold
is used. In all other cases,KFold
is used.Refer User Guide for the various cross-validation strategies that can be used here.
- refit : boolean, default=True
- Refit the best estimator with the entire dataset. If “False”, it is impossible to make predictions using this GridSearchCV instance after fitting.
- verbose : integer
- Controls the verbosity: the higher, the more messages.
- error_score : ‘raise’ (default) or numeric
- Value to assign to the score if an error occurs in estimator fitting. If set to ‘raise’, the error is raised. If a numeric value is given, FitFailedWarning is raised. This parameter does not affect the refit step, which will always raise the error.
- return_train_score : boolean, default=True
- If
'False'
, thecv_results_
attribute will not include training scores.
>>> from sklearn import svm, datasets >>> from sklearn.model_selection import GridSearchCV >>> iris = datasets.load_iris() >>> parameters = {'kernel':('linear', 'rbf'), 'C':[1, 10]} >>> svr = svm.SVC() >>> clf = GridSearchCV(svr, parameters) >>> clf.fit(iris.data, iris.target) ... GridSearchCV(cv=None, error_score=..., estimator=SVC(C=1.0, cache_size=..., class_weight=..., coef0=..., decision_function_shape=None, degree=..., gamma=..., kernel='rbf', max_iter=-1, probability=False, random_state=None, shrinking=True, tol=..., verbose=False), fit_params={}, iid=..., n_jobs=1, param_grid=..., pre_dispatch=..., refit=..., return_train_score=..., scoring=..., verbose=...) >>> sorted(clf.cv_results_.keys()) ... ['mean_fit_time', 'mean_score_time', 'mean_test_score',... 'mean_train_score', 'param_C', 'param_kernel', 'params',... 'rank_test_score', 'split0_test_score',... 'split0_train_score', 'split1_test_score', 'split1_train_score',... 'split2_test_score', 'split2_train_score',... 'std_fit_time', 'std_score_time', 'std_test_score', 'std_train_score'...]
- cv_results_ : dict of numpy (masked) ndarrays
A dict with keys as column headers and values as columns, that can be imported into a pandas
DataFrame
.For instance the below given table
param_kernel param_gamma param_degree split0_test_score ... rank_.... ‘poly’ – 2 0.8 ... 2 ‘poly’ – 3 0.7 ... 4 ‘rbf’ 0.1 – 0.8 ... 3 ‘rbf’ 0.2 – 0.9 ... 1 will be represented by a
cv_results_
dict of:{ 'param_kernel': masked_array(data = ['poly', 'poly', 'rbf', 'rbf'], mask = [False False False False]...) 'param_gamma': masked_array(data = [-- -- 0.1 0.2], mask = [ True True False False]...), 'param_degree': masked_array(data = [2.0 3.0 -- --], mask = [False False True True]...), 'split0_test_score' : [0.8, 0.7, 0.8, 0.9], 'split1_test_score' : [0.82, 0.5, 0.7, 0.78], 'mean_test_score' : [0.81, 0.60, 0.75, 0.82], 'std_test_score' : [0.02, 0.01, 0.03, 0.03], 'rank_test_score' : [2, 4, 3, 1], 'split0_train_score' : [0.8, 0.9, 0.7], 'split1_train_score' : [0.82, 0.5, 0.7], 'mean_train_score' : [0.81, 0.7, 0.7], 'std_train_score' : [0.03, 0.03, 0.04], 'mean_fit_time' : [0.73, 0.63, 0.43, 0.49], 'std_fit_time' : [0.01, 0.02, 0.01, 0.01], 'mean_score_time' : [0.007, 0.06, 0.04, 0.04], 'std_score_time' : [0.001, 0.002, 0.003, 0.005], 'params' : [{'kernel': 'poly', 'degree': 2}, ...], }
NOTE that the key
'params'
is used to store a list of parameter settings dict for all the parameter candidates.The
mean_fit_time
,std_fit_time
,mean_score_time
andstd_score_time
are all in seconds.- best_estimator_ : estimator
- Estimator that was chosen by the search, i.e. estimator which gave highest score (or smallest loss if specified) on the left out data. Not available if refit=False.
- best_score_ : float
- Score of best_estimator on the left out data.
- best_params_ : dict
- Parameter setting that gave the best results on the hold out data.
- best_index_ : int
The index (of the
cv_results_
arrays) which corresponds to the best candidate parameter setting.The dict at
search.cv_results_['params'][search.best_index_]
gives the parameter setting for the best model, that gives the highest mean score (search.best_score_
).- scorer_ : function
- Scorer function used on the held out data to choose the best parameters for the model.
- n_splits_ : int
- The number of cross-validation splits (folds/iterations).
The parameters selected are those that maximize the score of the left out data, unless an explicit score is passed in which case it is used instead.
If n_jobs was set to a value higher than one, the data is copied for each point in the grid (and not n_jobs times). This is done for efficiency reasons if individual jobs take very little time, but may raise errors if the dataset is large and not enough memory is available. A workaround in this case is to set pre_dispatch. Then, the memory is copied only pre_dispatch many times. A reasonable value for pre_dispatch is 2 * n_jobs.
ParameterGrid
:- generates all the combinations of a hyperparameter grid.
sklearn.model_selection.train_test_split()
:- utility function to split the data into a development set usable for fitting a GridSearchCV instance and an evaluation set for its final evaluation.
sklearn.metrics.make_scorer()
:- Make a scorer from a performance metric or loss function.
-
fit
(X, y=None, groups=None)¶ Run fit with all sets of parameters.
- X : array-like, shape = [n_samples, n_features]
- Training vector, where n_samples is the number of samples and n_features is the number of features.
- y : array-like, shape = [n_samples] or [n_samples, n_output], optional
- Target relative to X for classification or regression; None for unsupervised learning.
- groups : array-like, with shape (n_samples,), optional
- Group labels for the samples used while splitting the dataset into train/test set.
-
class
PREDICT.processing.SearchCV.
GridSearchCVfastr
(estimator, param_grid, scoring=None, fit_params=None, n_jobs=1, iid=True, refit=True, cv=None, verbose=0, pre_dispatch='2*n_jobs', error_score='raise', return_train_score=True)¶ Bases:
PREDICT.processing.SearchCV.BaseSearchCVfastr
Exhaustive search over specified parameter values for an estimator.
Important members are fit, predict.
GridSearchCV implements a “fit” and a “score” method. It also implements “predict”, “predict_proba”, “decision_function”, “transform” and “inverse_transform” if they are implemented in the estimator used.
The parameters of the estimator used to apply these methods are optimized by cross-validated grid-search over a parameter grid.
Read more in the User Guide.
- estimator : estimator object.
- This is assumed to implement the scikit-learn estimator interface.
Either estimator needs to provide a
score
function, orscoring
must be passed. - param_grid : dict or list of dictionaries
- Dictionary with parameters names (string) as keys and lists of parameter settings to try as values, or a list of such dictionaries, in which case the grids spanned by each dictionary in the list are explored. This enables searching over any sequence of parameter settings.
- scoring : string, callable or None, default=None
- A string (see model evaluation documentation) or
a scorer callable object / function with signature
scorer(estimator, X, y)
. IfNone
, thescore
method of the estimator is used. - fit_params : dict, optional
- Parameters to pass to the fit method.
- n_jobs : int, default=1
- Number of jobs to run in parallel.
- pre_dispatch : int, or string, optional
Controls the number of jobs that get dispatched during parallel execution. Reducing this number can be useful to avoid an explosion of memory consumption when more jobs get dispatched than CPUs can process. This parameter can be:
- None, in which case all the jobs are immediately created and spawned. Use this for lightweight and fast-running jobs, to avoid delays due to on-demand spawning of the jobs
- An int, giving the exact number of total jobs that are spawned
- A string, giving an expression as a function of n_jobs, as in ‘2*n_jobs’
- iid : boolean, default=True
- If True, the data is assumed to be identically distributed across the folds, and the loss minimized is the total loss per sample, and not the mean loss across the folds.
- cv : int, cross-validation generator or an iterable, optional
Determines the cross-validation splitting strategy. Possible inputs for cv are:
- None, to use the default 3-fold cross validation,
- integer, to specify the number of folds in a (Stratified)KFold,
- An object to be used as a cross-validation generator.
- An iterable yielding train, test splits.
For integer/None inputs, if the estimator is a classifier and
y
is either binary or multiclass,StratifiedKFold
is used. In all other cases,KFold
is used.Refer User Guide for the various cross-validation strategies that can be used here.
- refit : boolean, default=True
- Refit the best estimator with the entire dataset. If “False”, it is impossible to make predictions using this GridSearchCV instance after fitting.
- verbose : integer
- Controls the verbosity: the higher, the more messages.
- error_score : ‘raise’ (default) or numeric
- Value to assign to the score if an error occurs in estimator fitting. If set to ‘raise’, the error is raised. If a numeric value is given, FitFailedWarning is raised. This parameter does not affect the refit step, which will always raise the error.
- return_train_score : boolean, default=True
- If
'False'
, thecv_results_
attribute will not include training scores.
>>> from sklearn import svm, datasets >>> from sklearn.model_selection import GridSearchCV >>> iris = datasets.load_iris() >>> parameters = {'kernel':('linear', 'rbf'), 'C':[1, 10]} >>> svr = svm.SVC() >>> clf = GridSearchCV(svr, parameters) >>> clf.fit(iris.data, iris.target) ... GridSearchCV(cv=None, error_score=..., estimator=SVC(C=1.0, cache_size=..., class_weight=..., coef0=..., decision_function_shape=None, degree=..., gamma=..., kernel='rbf', max_iter=-1, probability=False, random_state=None, shrinking=True, tol=..., verbose=False), fit_params={}, iid=..., n_jobs=1, param_grid=..., pre_dispatch=..., refit=..., return_train_score=..., scoring=..., verbose=...) >>> sorted(clf.cv_results_.keys()) ... ['mean_fit_time', 'mean_score_time', 'mean_test_score',... 'mean_train_score', 'param_C', 'param_kernel', 'params',... 'rank_test_score', 'split0_test_score',... 'split0_train_score', 'split1_test_score', 'split1_train_score',... 'split2_test_score', 'split2_train_score',... 'std_fit_time', 'std_score_time', 'std_test_score', 'std_train_score'...]
- cv_results_ : dict of numpy (masked) ndarrays
A dict with keys as column headers and values as columns, that can be imported into a pandas
DataFrame
.For instance the below given table
param_kernel param_gamma param_degree split0_test_score ... rank_.... ‘poly’ – 2 0.8 ... 2 ‘poly’ – 3 0.7 ... 4 ‘rbf’ 0.1 – 0.8 ... 3 ‘rbf’ 0.2 – 0.9 ... 1 will be represented by a
cv_results_
dict of:{ 'param_kernel': masked_array(data = ['poly', 'poly', 'rbf', 'rbf'], mask = [False False False False]...) 'param_gamma': masked_array(data = [-- -- 0.1 0.2], mask = [ True True False False]...), 'param_degree': masked_array(data = [2.0 3.0 -- --], mask = [False False True True]...), 'split0_test_score' : [0.8, 0.7, 0.8, 0.9], 'split1_test_score' : [0.82, 0.5, 0.7, 0.78], 'mean_test_score' : [0.81, 0.60, 0.75, 0.82], 'std_test_score' : [0.02, 0.01, 0.03, 0.03], 'rank_test_score' : [2, 4, 3, 1], 'split0_train_score' : [0.8, 0.9, 0.7], 'split1_train_score' : [0.82, 0.5, 0.7], 'mean_train_score' : [0.81, 0.7, 0.7], 'std_train_score' : [0.03, 0.03, 0.04], 'mean_fit_time' : [0.73, 0.63, 0.43, 0.49], 'std_fit_time' : [0.01, 0.02, 0.01, 0.01], 'mean_score_time' : [0.007, 0.06, 0.04, 0.04], 'std_score_time' : [0.001, 0.002, 0.003, 0.005], 'params' : [{'kernel': 'poly', 'degree': 2}, ...], }
NOTE that the key
'params'
is used to store a list of parameter settings dict for all the parameter candidates.The
mean_fit_time
,std_fit_time
,mean_score_time
andstd_score_time
are all in seconds.- best_estimator_ : estimator
- Estimator that was chosen by the search, i.e. estimator which gave highest score (or smallest loss if specified) on the left out data. Not available if refit=False.
- best_score_ : float
- Score of best_estimator on the left out data.
- best_params_ : dict
- Parameter setting that gave the best results on the hold out data.
- best_index_ : int
The index (of the
cv_results_
arrays) which corresponds to the best candidate parameter setting.The dict at
search.cv_results_['params'][search.best_index_]
gives the parameter setting for the best model, that gives the highest mean score (search.best_score_
).- scorer_ : function
- Scorer function used on the held out data to choose the best parameters for the model.
- n_splits_ : int
- The number of cross-validation splits (folds/iterations).
The parameters selected are those that maximize the score of the left out data, unless an explicit score is passed in which case it is used instead.
If n_jobs was set to a value higher than one, the data is copied for each point in the grid (and not n_jobs times). This is done for efficiency reasons if individual jobs take very little time, but may raise errors if the dataset is large and not enough memory is available. A workaround in this case is to set pre_dispatch. Then, the memory is copied only pre_dispatch many times. A reasonable value for pre_dispatch is 2 * n_jobs.
ParameterGrid
:- generates all the combinations of a hyperparameter grid.
sklearn.model_selection.train_test_split()
:- utility function to split the data into a development set usable for fitting a GridSearchCV instance and an evaluation set for its final evaluation.
sklearn.metrics.make_scorer()
:- Make a scorer from a performance metric or loss function.
-
fit
(X, y=None, groups=None)¶ Run fit with all sets of parameters.
- X : array-like, shape = [n_samples, n_features]
- Training vector, where n_samples is the number of samples and n_features is the number of features.
- y : array-like, shape = [n_samples] or [n_samples, n_output], optional
- Target relative to X for classification or regression; None for unsupervised learning.
- groups : array-like, with shape (n_samples,), optional
- Group labels for the samples used while splitting the dataset into train/test set.
-
class
PREDICT.processing.SearchCV.
RandomizedSearchCVJoblib
(estimator, param_distributions={}, n_iter=10, scoring=None, fit_params=None, n_jobs=1, iid=True, refit=True, cv=None, verbose=0, pre_dispatch='2*n_jobs', random_state=None, error_score='raise', return_train_score=True, n_jobspercore=100)¶ Bases:
PREDICT.processing.SearchCV.BaseSearchCVJoblib
Randomized search on hyper parameters.
RandomizedSearchCV implements a “fit” and a “score” method. It also implements “predict”, “predict_proba”, “decision_function”, “transform” and “inverse_transform” if they are implemented in the estimator used.
The parameters of the estimator used to apply these methods are optimized by cross-validated search over parameter settings.
In contrast to GridSearchCV, not all parameter values are tried out, but rather a fixed number of parameter settings is sampled from the specified distributions. The number of parameter settings that are tried is given by n_iter.
If all parameters are presented as a list, sampling without replacement is performed. If at least one parameter is given as a distribution, sampling with replacement is used. It is highly recommended to use continuous distributions for continuous parameters.
Read more in the User Guide.
- estimator : estimator object.
- A object of that type is instantiated for each grid point.
This is assumed to implement the scikit-learn estimator interface.
Either estimator needs to provide a
score
function, orscoring
must be passed. - param_distributions : dict
- Dictionary with parameters names (string) as keys and distributions
or lists of parameters to try. Distributions must provide a
rvs
method for sampling (such as those from scipy.stats.distributions). If a list is given, it is sampled uniformly. - n_iter : int, default=10
- Number of parameter settings that are sampled. n_iter trades off runtime vs quality of the solution.
- scoring : string, callable or None, default=None
- A string (see model evaluation documentation) or
a scorer callable object / function with signature
scorer(estimator, X, y)
. IfNone
, thescore
method of the estimator is used. - fit_params : dict, optional
- Parameters to pass to the fit method.
- n_jobs : int, default=1
- Number of jobs to run in parallel.
- pre_dispatch : int, or string, optional
Controls the number of jobs that get dispatched during parallel execution. Reducing this number can be useful to avoid an explosion of memory consumption when more jobs get dispatched than CPUs can process. This parameter can be:
- None, in which case all the jobs are immediately created and spawned. Use this for lightweight and fast-running jobs, to avoid delays due to on-demand spawning of the jobs
- An int, giving the exact number of total jobs that are spawned
- A string, giving an expression as a function of n_jobs, as in ‘2*n_jobs’
- iid : boolean, default=True
- If True, the data is assumed to be identically distributed across the folds, and the loss minimized is the total loss per sample, and not the mean loss across the folds.
- cv : int, cross-validation generator or an iterable, optional
Determines the cross-validation splitting strategy. Possible inputs for cv are:
- None, to use the default 3-fold cross validation,
- integer, to specify the number of folds in a (Stratified)KFold,
- An object to be used as a cross-validation generator.
- An iterable yielding train, test splits.
For integer/None inputs, if the estimator is a classifier and
y
is either binary or multiclass,StratifiedKFold
is used. In all other cases,KFold
is used.Refer User Guide for the various cross-validation strategies that can be used here.
- refit : boolean, default=True
- Refit the best estimator with the entire dataset. If “False”, it is impossible to make predictions using this RandomizedSearchCV instance after fitting.
- verbose : integer
- Controls the verbosity: the higher, the more messages.
- random_state : int or RandomState
- Pseudo random number generator state used for random uniform sampling from lists of possible values instead of scipy.stats distributions.
- error_score : ‘raise’ (default) or numeric
- Value to assign to the score if an error occurs in estimator fitting. If set to ‘raise’, the error is raised. If a numeric value is given, FitFailedWarning is raised. This parameter does not affect the refit step, which will always raise the error.
- return_train_score : boolean, default=True
- If
'False'
, thecv_results_
attribute will not include training scores.
- cv_results_ : dict of numpy (masked) ndarrays
A dict with keys as column headers and values as columns, that can be imported into a pandas
DataFrame
.For instance the below given table
param_kernel param_gamma split0_test_score ... rank_test_score ‘rbf’ 0.1 0.8 ... 2 ‘rbf’ 0.2 0.9 ... 1 ‘rbf’ 0.3 0.7 ... 1 will be represented by a
cv_results_
dict of:{ 'param_kernel' : masked_array(data = ['rbf', 'rbf', 'rbf'], mask = False), 'param_gamma' : masked_array(data = [0.1 0.2 0.3], mask = False), 'split0_test_score' : [0.8, 0.9, 0.7], 'split1_test_score' : [0.82, 0.5, 0.7], 'mean_test_score' : [0.81, 0.7, 0.7], 'std_test_score' : [0.02, 0.2, 0.], 'rank_test_score' : [3, 1, 1], 'split0_train_score' : [0.8, 0.9, 0.7], 'split1_train_score' : [0.82, 0.5, 0.7], 'mean_train_score' : [0.81, 0.7, 0.7], 'std_train_score' : [0.03, 0.03, 0.04], 'mean_fit_time' : [0.73, 0.63, 0.43, 0.49], 'std_fit_time' : [0.01, 0.02, 0.01, 0.01], 'mean_score_time' : [0.007, 0.06, 0.04, 0.04], 'std_score_time' : [0.001, 0.002, 0.003, 0.005], 'params' : [{'kernel' : 'rbf', 'gamma' : 0.1}, ...], }
NOTE that the key
'params'
is used to store a list of parameter settings dict for all the parameter candidates.The
mean_fit_time
,std_fit_time
,mean_score_time
andstd_score_time
are all in seconds.- best_estimator_ : estimator
- Estimator that was chosen by the search, i.e. estimator which gave highest score (or smallest loss if specified) on the left out data. Not available if refit=False.
- best_score_ : float
- Score of best_estimator on the left out data.
- best_params_ : dict
- Parameter setting that gave the best results on the hold out data.
- best_index_ : int
The index (of the
cv_results_
arrays) which corresponds to the best candidate parameter setting.The dict at
search.cv_results_['params'][search.best_index_]
gives the parameter setting for the best model, that gives the highest mean score (search.best_score_
).- scorer_ : function
- Scorer function used on the held out data to choose the best parameters for the model.
- n_splits_ : int
- The number of cross-validation splits (folds/iterations).
The parameters selected are those that maximize the score of the held-out data, according to the scoring parameter.
If n_jobs was set to a value higher than one, the data is copied for each parameter setting(and not n_jobs times). This is done for efficiency reasons if individual jobs take very little time, but may raise errors if the dataset is large and not enough memory is available. A workaround in this case is to set pre_dispatch. Then, the memory is copied only pre_dispatch many times. A reasonable value for pre_dispatch is 2 * n_jobs.
GridSearchCV
:- Does exhaustive search over a grid of parameters.
ParameterSampler
:- A generator over parameter settins, constructed from param_distributions.
-
fit
(X, y=None, groups=None)¶ Run fit on the estimator with randomly drawn parameters.
- X : array-like, shape = [n_samples, n_features]
- Training vector, where n_samples in the number of samples and n_features is the number of features.
- y : array-like, shape = [n_samples] or [n_samples, n_output], optional
- Target relative to X for classification or regression; None for unsupervised learning.
- groups : array-like, with shape (n_samples,), optional
- Group labels for the samples used while splitting the dataset into train/test set.
-
class
PREDICT.processing.SearchCV.
RandomizedSearchCVfastr
(estimator, param_distributions={}, n_iter=10, scoring=None, fit_params=None, n_jobs=1, iid=True, refit=True, cv=None, verbose=0, pre_dispatch='2*n_jobs', random_state=None, error_score='raise', return_train_score=True, n_jobspercore=100, fastr_plugin=None)¶ Bases:
PREDICT.processing.SearchCV.BaseSearchCVfastr
Randomized search on hyper parameters.
RandomizedSearchCV implements a “fit” and a “score” method. It also implements “predict”, “predict_proba”, “decision_function”, “transform” and “inverse_transform” if they are implemented in the estimator used.
The parameters of the estimator used to apply these methods are optimized by cross-validated search over parameter settings.
In contrast to GridSearchCV, not all parameter values are tried out, but rather a fixed number of parameter settings is sampled from the specified distributions. The number of parameter settings that are tried is given by n_iter.
If all parameters are presented as a list, sampling without replacement is performed. If at least one parameter is given as a distribution, sampling with replacement is used. It is highly recommended to use continuous distributions for continuous parameters.
Read more in the User Guide.
- estimator : estimator object.
- A object of that type is instantiated for each grid point.
This is assumed to implement the scikit-learn estimator interface.
Either estimator needs to provide a
score
function, orscoring
must be passed. - param_distributions : dict
- Dictionary with parameters names (string) as keys and distributions
or lists of parameters to try. Distributions must provide a
rvs
method for sampling (such as those from scipy.stats.distributions). If a list is given, it is sampled uniformly. - n_iter : int, default=10
- Number of parameter settings that are sampled. n_iter trades off runtime vs quality of the solution.
- scoring : string, callable or None, default=None
- A string (see model evaluation documentation) or
a scorer callable object / function with signature
scorer(estimator, X, y)
. IfNone
, thescore
method of the estimator is used. - fit_params : dict, optional
- Parameters to pass to the fit method.
- n_jobs : int, default=1
- Number of jobs to run in parallel.
- pre_dispatch : int, or string, optional
Controls the number of jobs that get dispatched during parallel execution. Reducing this number can be useful to avoid an explosion of memory consumption when more jobs get dispatched than CPUs can process. This parameter can be:
- None, in which case all the jobs are immediately created and spawned. Use this for lightweight and fast-running jobs, to avoid delays due to on-demand spawning of the jobs
- An int, giving the exact number of total jobs that are spawned
- A string, giving an expression as a function of n_jobs, as in ‘2*n_jobs’
- iid : boolean, default=True
- If True, the data is assumed to be identically distributed across the folds, and the loss minimized is the total loss per sample, and not the mean loss across the folds.
- cv : int, cross-validation generator or an iterable, optional
Determines the cross-validation splitting strategy. Possible inputs for cv are:
- None, to use the default 3-fold cross validation,
- integer, to specify the number of folds in a (Stratified)KFold,
- An object to be used as a cross-validation generator.
- An iterable yielding train, test splits.
For integer/None inputs, if the estimator is a classifier and
y
is either binary or multiclass,StratifiedKFold
is used. In all other cases,KFold
is used.Refer User Guide for the various cross-validation strategies that can be used here.
- refit : boolean, default=True
- Refit the best estimator with the entire dataset. If “False”, it is impossible to make predictions using this RandomizedSearchCV instance after fitting.
- verbose : integer
- Controls the verbosity: the higher, the more messages.
- random_state : int or RandomState
- Pseudo random number generator state used for random uniform sampling from lists of possible values instead of scipy.stats distributions.
- error_score : ‘raise’ (default) or numeric
- Value to assign to the score if an error occurs in estimator fitting. If set to ‘raise’, the error is raised. If a numeric value is given, FitFailedWarning is raised. This parameter does not affect the refit step, which will always raise the error.
- return_train_score : boolean, default=True
- If
'False'
, thecv_results_
attribute will not include training scores.
- cv_results_ : dict of numpy (masked) ndarrays
A dict with keys as column headers and values as columns, that can be imported into a pandas
DataFrame
.For instance the below given table
param_kernel param_gamma split0_test_score ... rank_test_score ‘rbf’ 0.1 0.8 ... 2 ‘rbf’ 0.2 0.9 ... 1 ‘rbf’ 0.3 0.7 ... 1 will be represented by a
cv_results_
dict of:{ 'param_kernel' : masked_array(data = ['rbf', 'rbf', 'rbf'], mask = False), 'param_gamma' : masked_array(data = [0.1 0.2 0.3], mask = False), 'split0_test_score' : [0.8, 0.9, 0.7], 'split1_test_score' : [0.82, 0.5, 0.7], 'mean_test_score' : [0.81, 0.7, 0.7], 'std_test_score' : [0.02, 0.2, 0.], 'rank_test_score' : [3, 1, 1], 'split0_train_score' : [0.8, 0.9, 0.7], 'split1_train_score' : [0.82, 0.5, 0.7], 'mean_train_score' : [0.81, 0.7, 0.7], 'std_train_score' : [0.03, 0.03, 0.04], 'mean_fit_time' : [0.73, 0.63, 0.43, 0.49], 'std_fit_time' : [0.01, 0.02, 0.01, 0.01], 'mean_score_time' : [0.007, 0.06, 0.04, 0.04], 'std_score_time' : [0.001, 0.002, 0.003, 0.005], 'params' : [{'kernel' : 'rbf', 'gamma' : 0.1}, ...], }
NOTE that the key
'params'
is used to store a list of parameter settings dict for all the parameter candidates.The
mean_fit_time
,std_fit_time
,mean_score_time
andstd_score_time
are all in seconds.- best_estimator_ : estimator
- Estimator that was chosen by the search, i.e. estimator which gave highest score (or smallest loss if specified) on the left out data. Not available if refit=False.
- best_score_ : float
- Score of best_estimator on the left out data.
- best_params_ : dict
- Parameter setting that gave the best results on the hold out data.
- best_index_ : int
The index (of the
cv_results_
arrays) which corresponds to the best candidate parameter setting.The dict at
search.cv_results_['params'][search.best_index_]
gives the parameter setting for the best model, that gives the highest mean score (search.best_score_
).- scorer_ : function
- Scorer function used on the held out data to choose the best parameters for the model.
- n_splits_ : int
- The number of cross-validation splits (folds/iterations).
The parameters selected are those that maximize the score of the held-out data, according to the scoring parameter.
If n_jobs was set to a value higher than one, the data is copied for each parameter setting(and not n_jobs times). This is done for efficiency reasons if individual jobs take very little time, but may raise errors if the dataset is large and not enough memory is available. A workaround in this case is to set pre_dispatch. Then, the memory is copied only pre_dispatch many times. A reasonable value for pre_dispatch is 2 * n_jobs.
GridSearchCV
:- Does exhaustive search over a grid of parameters.
ParameterSampler
:- A generator over parameter settings, constructed from param_distributions.
-
fit
(X, y=None, groups=None)¶ Run fit on the estimator with randomly drawn parameters.
- X : array-like, shape = [n_samples, n_features]
- Training vector, where n_samples in the number of samples and n_features is the number of features.
- y : array-like, shape = [n_samples] or [n_samples, n_output], optional
- Target relative to X for classification or regression; None for unsupervised learning.
- groups : array-like, with shape (n_samples,), optional
- Group labels for the samples used while splitting the dataset into train/test set.
-
PREDICT.processing.SearchCV.
chunks
(l, n)¶ Yield successive n-sized chunks from l.
-
PREDICT.processing.SearchCV.
chunksdict
(data, SIZE)¶ Split a dictionary in equal parts of certain slice
-
PREDICT.processing.SearchCV.
rms_score
(truth, prediction)¶ Root-mean-square-error metric
-
PREDICT.processing.SearchCV.
sar_score
(truth, prediction)¶ SAR metric from Caruana et al. 2004
PREDICT.processing.fitandscore module¶
-
PREDICT.processing.fitandscore.
delete_nonestimator_parameters
(parameters)¶ Delete all parameters in a parameter dictionary that are not used for the actual estimator.
-
PREDICT.processing.fitandscore.
fit_and_score
(estimator, X, y, scorer, train, test, para, fit_params=None, return_train_score=True, return_n_test_samples=True, return_times=True, return_parameters=True, error_score='raise', verbose=True, return_all=True)¶ Fit an estimator to a dataset and score the performance. The following methods can currently be applied as preprocessing before fitting, in this order: 1. Select features based on feature type group (e.g. shape, histogram). 2. Apply feature imputation (WIP). 3. Apply feature selection based on variance of feature among patients. 4. Univariate statistical testing (e.g. t-test, Wilcoxon). 5. Scale features with e.g. z-scoring. 6. Use Relief feature selection. 7. Select features based on a fit with a LASSO model. 8. Select features using PCA. 9. If a SingleLabel classifier is used for a MultiLabel problem,
a OneVsRestClassifier is employed around it.All of the steps are optional.
- estimator: sklearn estimator, mandatory
- Unfitted estimator which will be fit.
- X: array, mandatory
- Array containingfor each object (rows) the feature values (1st Column) and the associated feature label (2nd Column).
- y: list(?), mandatory
- List containing the labels of the objects.
- scorer: sklearn scorer, mandatory
- Function used as optimization criterion for the hyperparamater optimization.
- train: list, mandatory
- Indices of the objects to be used as training set.
- test: list, mandatory
- Indices of the objects to be used as testing set.
- para: dictionary, mandatory
- Contains the settings used for the above preprocessing functions and the fitting. TODO: Create a default object and show the fields.
- fit_params:dictionary, default None
- Parameters supplied to the estimator for fitting. See the SKlearn site for the parameters of the estimators.
- return_train_score: boolean, default True
- Save the training score to the final SearchCV object.
- return_n_test_samples: boolean, default True
- Save the number of times each sample was used in the test set to the final SearchCV object.
- return_times: boolean, default True
- Save the time spend for each fit to the final SearchCV object.
- return_parameters: boolean, default True
- Return the parameters used in the final fit to the final SearchCV object.
- error_score: numeric or “raise” by default
- Value to assign to the score if an error occurs in estimator fitting. If set to “raise”, the error is raised. If a numeric value is given, FitFailedWarning is raised. This parameter does not affect the refit step, which will always raise the error.
- verbose: boolean, default=True
- If True, print intermediate progress to command line. Warnings are always printed.
- return_all: boolean, default=True
- If False, only the ret object containing the performance will be returned. If True, the ret object plus all fitted objects will be returned.
Depending on the return_all input parameter, either only ret or all objects below are returned.
- ret: list
- Contains optionally the train_scores and the test_scores, test_sample_counts, fit_time, score_time, parameters_est and parameters_all.
- GroupSel: PREDICT GroupSel Object
- Either None if the GroupSelFitted GroupSel Object.
- VarSel: PREDICT GroupSel Object
- Either None if the GroupSelFitted GroupSel Object.
- SelectModel: PREDICT GroupSel Object
- Either None if the GroupSelFitted GroupSel Object.
feature_labels
scaler
- imputer: PREDICT GroupSel Object
- Either None if the GroupSelFitted GroupSel Object.
- pca: PREDICT GroupSel Object
- Either None if the GroupSelFitted GroupSel Object.
- StatisticalSel: PREDICT GroupSel Object
- Either None if the GroupSelFitted GroupSel Object.
- ReliefSel: PREDICT GroupSel Object
- Either None if the GroupSelFitted GroupSel Object.
-
PREDICT.processing.fitandscore.
replacenan
(image_features, verbose=True, feature_labels=None)¶ Replace the NaNs in an image feature matrix.