GridSearchCV

task:

Search

subtask:

grid

host:

sklearn

function:

GridSearchCV

input tokens (receivers):

dfy : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

dfx : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

estimator : instance of a machine learning class

types: (“<type ‘str’>”, “<class ‘sklearn.linear_model.base.LinearRegression’>”, “<class ‘sklearn.linear_model.ridge.Ridge’>”, “<class ‘sklearn.kernel_ridge.KernelRidge’>”, “<class ‘sklearn.linear_model.coordinate_descent.Lasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskLasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.ElasticNet’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskElasticNet’>”, “<class ‘sklearn.linear_model.least_angle.Lars’>”, “<class ‘sklearn.linear_model.least_angle.LassoLars’>”, “<class ‘sklearn.linear_model.bayes.BayesianRidge’>”, “<class ‘sklearn.linear_model.bayes.ARDRegression’>”, “<class ‘sklearn.linear_model.logistic.LogisticRegression’>”, “<class ‘sklearn.linear_model.stochastic_gradient.SGDRegressor’>”, “<class ‘sklearn.svm.classes.SVR’>”, “<class ‘sklearn.svm.classes.NuSVR’>”, “<class ‘sklearn.svm.classes.LinearSVR’>”, “<class ‘sklearn.neural_network.multilayer_perceptron.MLPRegressor’>”, “<class ‘chemml.nn.keras.mlp.MLP_sklearn’>”)

scorer : instance of scikit-learn’s make_scorer class

types: (“<class ‘sklearn.metrics.scorer._PredictScorer’>”,)

cv : instance of scikit-learn’s cross validation generator or instance object

types: (“<type ‘generator’>”, “<class ‘sklearn.model_selection._split.KFold’>”, “<class ‘sklearn.model_selection._split.ShuffleSplit’>”, “<class ‘sklearn.model_selection._split.StratifiedShuffleSplit’>”, “<class ‘sklearn.model_selection._split.LeaveOneOut’>”)

output tokens (senders):

cv_results_ : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

api : instance of scikit-learn’s GridSearchCV class

types: (“<class ‘sklearn.grid_search.GridSearchCV’>”,)

best_estimator_ : instance of a machine learning class

types: (“<class ‘sklearn.linear_model.base.LinearRegression’>”, “<class ‘sklearn.linear_model.ridge.Ridge’>”, “<class ‘sklearn.kernel_ridge.KernelRidge’>”, “<class ‘sklearn.linear_model.coordinate_descent.Lasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskLasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.ElasticNet’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskElasticNet’>”, “<class ‘sklearn.linear_model.least_angle.Lars’>”, “<class ‘sklearn.linear_model.least_angle.LassoLars’>”, “<class ‘sklearn.linear_model.bayes.BayesianRidge’>”, “<class ‘sklearn.linear_model.bayes.ARDRegression’>”, “<class ‘sklearn.linear_model.logistic.LogisticRegression’>”, “<class ‘sklearn.linear_model.stochastic_gradient.SGDRegressor’>”, “<class ‘sklearn.svm.classes.SVR’>”, “<class ‘sklearn.svm.classes.NuSVR’>”, “<class ‘sklearn.svm.classes.LinearSVR’>”, “<class ‘sklearn.neural_network.multilayer_perceptron.MLPRegressor’>”, “<class ‘chemml.nn.keras.mlp.MLP_sklearn’>”)

wrapper parameters:

track_header : Boolean, (default:True)

if True, the input dataframe’s header will be transformed to the output dataframe

choose one of: (True, False)

required packages:

scikit-learn, 0.19.0

pandas, 0.20.3

config file view:

##

<< host = sklearn    << function = GridSearchCV

<< track_header = True

<< scoring = None

<< n_jobs = 1

<< verbose = 0

<< fit_params = None

<< refit = True

<< return_train_score = True

<< iid = True

<< estimator = @estimator

<< error_score = raise

<< pre_dispatch = 2 * n_jobs

<< param_grid = {}

<< cv = None

>> id dfy

>> id dfx

>> id estimator

>> id scorer

>> id cv

>> id cv_results_

>> id api

>> id best_estimator_

Note

The documentation page for function parameters: http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.html

cross_val_predict

task:

Search

subtask:

validate

host:

sklearn

function:

cross_val_predict

input tokens (receivers):

dfy : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

dfx : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

estimator : instance of a machine learning class

types: (“<class ‘sklearn.linear_model.base.LinearRegression’>”, “<class ‘sklearn.linear_model.ridge.Ridge’>”, “<class ‘sklearn.kernel_ridge.KernelRidge’>”, “<class ‘sklearn.linear_model.coordinate_descent.Lasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskLasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.ElasticNet’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskElasticNet’>”, “<class ‘sklearn.linear_model.least_angle.Lars’>”, “<class ‘sklearn.linear_model.least_angle.LassoLars’>”, “<class ‘sklearn.linear_model.bayes.BayesianRidge’>”, “<class ‘sklearn.linear_model.bayes.ARDRegression’>”, “<class ‘sklearn.linear_model.logistic.LogisticRegression’>”, “<class ‘sklearn.linear_model.stochastic_gradient.SGDRegressor’>”, “<class ‘sklearn.svm.classes.SVR’>”, “<class ‘sklearn.svm.classes.NuSVR’>”, “<class ‘sklearn.svm.classes.LinearSVR’>”, “<class ‘sklearn.neural_network.multilayer_perceptron.MLPRegressor’>”, “<class ‘chemml.nn.keras.mlp.MLP_sklearn’>”)

scorer : instance of scikit-learn’s make_scorer class

types: (“<class ‘sklearn.metrics.scorer._PredictScorer’>”,)

cv : cross-validation generator or instance object

types: (“<type ‘generator’>”, “<class ‘sklearn.model_selection._split.KFold’>”, “<class ‘sklearn.model_selection._split.ShuffleSplit’>”, “<class ‘sklearn.model_selection._split.StratifiedShuffleSplit’>”, “<class ‘sklearn.model_selection._split.LeaveOneOut’>”)

output tokens (senders):

dfy_predict : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

wrapper parameters:

track_header : Boolean, (default:True)

if True, the input dataframe’s header will be transformed to the output dataframe

choose one of: (True, False)

required packages:

scikit-learn, 0.19.0

pandas, 0.20.3

config file view:

##

<< host = sklearn    << function = cross_val_predict

<< track_header = True

<< n_jobs = 1

<< verbose = 0

<< fit_params = None

<< method = predict

<< pre_dispatch = 2 * n_jobs

<< estimator = @estimator

<< groups = None

<< y = None

<< X = @dfx

<< cv = None

>> id dfy

>> id dfx

>> id estimator

>> id scorer

>> id cv

>> id dfy_predict

Note

The documentation page for function parameters: http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.cross_val_predict.html#sklearn.model_selection.cross_val_predict

cross_val_score

task:

Search

subtask:

validate

host:

sklearn

function:

cross_val_score

input tokens (receivers):

dfy : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

dfx : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

estimator : instance of a machine learning class

types: (“<class ‘sklearn.linear_model.base.LinearRegression’>”, “<class ‘sklearn.linear_model.ridge.Ridge’>”, “<class ‘sklearn.kernel_ridge.KernelRidge’>”, “<class ‘sklearn.linear_model.coordinate_descent.Lasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskLasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.ElasticNet’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskElasticNet’>”, “<class ‘sklearn.linear_model.least_angle.Lars’>”, “<class ‘sklearn.linear_model.least_angle.LassoLars’>”, “<class ‘sklearn.linear_model.bayes.BayesianRidge’>”, “<class ‘sklearn.linear_model.bayes.ARDRegression’>”, “<class ‘sklearn.linear_model.logistic.LogisticRegression’>”, “<class ‘sklearn.linear_model.stochastic_gradient.SGDRegressor’>”, “<class ‘sklearn.svm.classes.SVR’>”, “<class ‘sklearn.svm.classes.NuSVR’>”, “<class ‘sklearn.svm.classes.LinearSVR’>”, “<class ‘sklearn.neural_network.multilayer_perceptron.MLPRegressor’>”, “<class ‘chemml.nn.keras.mlp.MLP_sklearn’>”)

scorer : instance of scikit-learn’s make_scorer class

types: (“<class ‘sklearn.metrics.scorer._PredictScorer’>”,)

cv : cross-validation generator or instance object

types: (“<type ‘generator’>”, “<class ‘sklearn.model_selection._split.KFold’>”, “<class ‘sklearn.model_selection._split.ShuffleSplit’>”, “<class ‘sklearn.model_selection._split.StratifiedShuffleSplit’>”, “<class ‘sklearn.model_selection._split.LeaveOneOut’>”)

output tokens (senders):

scores : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

wrapper parameters:

track_header : Boolean, (default:True)

if True, the input dataframe’s header will be transformed to the output dataframe

choose one of: (True, False)

required packages:

scikit-learn, 0.19.0

pandas, 0.20.3

config file view:

##

<< host = sklearn    << function = cross_val_score

<< track_header = True

<< scoring = None

<< n_jobs = 1

<< verbose = 0

<< fit_params = None

<< pre_dispatch = 2 * n_jobs

<< estimator = @estimator

<< groups = None

<< y = None

<< X = @dfx

<< cv = None

>> id dfy

>> id dfx

>> id estimator

>> id scorer

>> id cv

>> id scores

Note

The documentation page for function parameters: http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.cross_val_score.html#sklearn.model_selection.cross_val_score

evaluate_regression

task:

Search

subtask:

evaluate

host:

sklearn

function:

evaluate_regression

input tokens (receivers):

dfy : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

dfy_predict : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

output tokens (senders):

evaluation_results_ : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

evaluator : dictionary of metrics and their score function

types: (“<type ‘dict’>”,)

wrapper parameters:

mae_multioutput : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_absolute_error.html#sklearn.metrics.mean_absolute_error, (default:uniform_average)

choose one of: (‘raw_values’, ‘uniform_average’)

r2_score : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.r2_score.html#sklearn.metrics.r2_score, (default:False)

choose one of: (True, False)

mean_absolute_error : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_absolute_error.html#sklearn.metrics.mean_absolute_error, (default:False)

choose one of: (True, False)

multioutput : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.r2_score.html#sklearn.metrics.r2_score, (default:uniform_average)

choose one of: (‘raw_values’, ‘uniform_average’, ‘variance_weighted’)

mse_sample_weight : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html#sklearn.metrics.mean_squared_error, (default:None)

choose one of: []

rmse_multioutput : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html#sklearn.metrics.mean_squared_error, (default:uniform_average)

choose one of: (‘raw_values’, ‘uniform_average’)

median_absolute_error : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.median_absolute_error.html#sklearn.metrics.median_absolute_error, (default:False)

choose one of: (True, False)

mae_sample_weight : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_absolute_error.html#sklearn.metrics.mean_absolute_error, (default:None)

choose one of: []

rmse_sample_weight : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html#sklearn.metrics.mean_squared_error, (default:None)

choose one of: []

track_header : Boolean, (default:True)

if True, the input dataframe’s header will be transformed to the output dataframe

choose one of: (True, False)

mean_squared_error : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html#sklearn.metrics.mean_squared_error, (default:False)

choose one of: (True, False)

root_mean_squared_error : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html#sklearn.metrics.mean_squared_error, (default:False)

choose one of: (True, False)

explained_variance_score : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.explained_variance_score.html#sklearn.metrics.explained_variance_score, (default:False)

choose one of: (True, False)

r2_sample_weight : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.r2_score.html#sklearn.metrics.r2_score, (default:None)

choose one of: []

ev_sample_weight : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.explained_variance_score.html#sklearn.metrics.explained_variance_score, (default:None)

choose one of: []

ev_multioutput : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.explained_variance_score.html#sklearn.metrics.explained_variance_score, (default:uniform_average)

choose one of: (‘raw_values’, ‘uniform_average’, ‘variance_weighted’)

mse_multioutput : http://scikit-learn.org/stable/modules/generated/sklearn.metrics.mean_squared_error.html#sklearn.metrics.mean_squared_error, (default:uniform_average)

choose one of: (‘raw_values’, ‘uniform_average’)

required packages:

scikit-learn, 0.19.0

pandas, 0.20.3

config file view:

##

<< host = sklearn    << function = evaluate_regression

<< mae_multioutput = uniform_average

<< r2_score = False

<< mean_absolute_error = False

<< multioutput = uniform_average

<< mse_sample_weight = None

<< rmse_multioutput = uniform_average

<< median_absolute_error = False

<< mae_sample_weight = None

<< rmse_sample_weight = None

<< track_header = True

<< mean_squared_error = False

<< root_mean_squared_error = False

<< explained_variance_score = False

<< r2_sample_weight = None

<< ev_sample_weight = None

<< ev_multioutput = uniform_average

<< mse_multioutput = uniform_average

>> id dfy

>> id dfy_predict

>> id evaluation_results_

>> id evaluator

Note

The documentation page for function parameters: http://scikit-learn.org/dev/modules/model_evaluation.html#regression-metrics

learning_curve

task:

Search

subtask:

grid

host:

sklearn

function:

learning_curve

input tokens (receivers):

dfy : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

dfx : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

estimator : instance of a machine learning class

types: (“<class ‘sklearn.linear_model.base.LinearRegression’>”, “<class ‘sklearn.linear_model.ridge.Ridge’>”, “<class ‘sklearn.kernel_ridge.KernelRidge’>”, “<class ‘sklearn.linear_model.coordinate_descent.Lasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskLasso’>”, “<class ‘sklearn.linear_model.coordinate_descent.ElasticNet’>”, “<class ‘sklearn.linear_model.coordinate_descent.MultiTaskElasticNet’>”, “<class ‘sklearn.linear_model.least_angle.Lars’>”, “<class ‘sklearn.linear_model.least_angle.LassoLars’>”, “<class ‘sklearn.linear_model.bayes.BayesianRidge’>”, “<class ‘sklearn.linear_model.bayes.ARDRegression’>”, “<class ‘sklearn.linear_model.logistic.LogisticRegression’>”, “<class ‘sklearn.linear_model.stochastic_gradient.SGDRegressor’>”, “<class ‘sklearn.svm.classes.SVR’>”, “<class ‘sklearn.svm.classes.NuSVR’>”, “<class ‘sklearn.svm.classes.LinearSVR’>”, “<class ‘sklearn.neural_network.multilayer_perceptron.MLPRegressor’>”, “<class ‘chemml.nn.keras.mlp.MLP_sklearn’>”)

scorer : instance of scikit-learn’s make_scorer class

types: (“<class ‘sklearn.metrics.scorer._PredictScorer’>”,)

cv : instance of scikit-learn’s cross validation generator or instance object

types: (“<type ‘generator’>”, “<class ‘sklearn.model_selection._split.KFold’>”, “<class ‘sklearn.model_selection._split.ShuffleSplit’>”, “<class ‘sklearn.model_selection._split.StratifiedShuffleSplit’>”, “<class ‘sklearn.model_selection._split.LeaveOneOut’>”)

output tokens (senders):

train_sizes_abs : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

extended_result_ : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

test_scores : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

train_scores : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

wrapper parameters:

track_header : Boolean, (default:True)

if True, the input dataframe’s header will be transformed to the output dataframe

choose one of: (True, False)

required packages:

scikit-learn, 0.19.0

pandas, 0.20.3

config file view:

##

<< host = sklearn    << function = learning_curve

<< track_header = True

<< scoring = None

<< n_jobs = 1

<< shuffle = False

<< groups = None

<< random_state = None

<< pre_dispatch = all

<< estimator = @estimator

<< exploit_incremental_learning = False

<< train_sizes = [0.1, 0.33, 0.55, 0.78, 1.0]

<< y = None

<< X = @dfx

<< cv = None

<< verbose = 0

>> id dfy

>> id dfx

>> id estimator

>> id scorer

>> id cv

>> id train_sizes_abs

>> id extended_result_

>> id test_scores

>> id train_scores

Note

The documentation page for function parameters: http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.learning_curve.html#sklearn.model_selection.learning_curve

scorer_regression

task:

Search

subtask:

evaluate

host:

sklearn

function:

scorer_regression

input tokens (receivers):

this block doesn’t receive anything

output tokens (senders):

scorer : Callable object that returns a scalar score

types: (“<class ‘sklearn.metrics.scorer._PredictScorer’>”,)

wrapper parameters:

track_header : Boolean, (default:True)

if True, the input dataframe’s header will be transformed to the output dataframe

choose one of: (True, False)

metric : string: ‘mae’, ‘mse’, ‘r2’, (default:mae)

http://scikit-learn.org/dev/modules/model_evaluation.html#regression-metrics

choose one of: (‘mae’, ‘mse’, ‘r2’)

required packages:

scikit-learn, 0.19.0

pandas, 0.20.3

config file view:

##

<< host = sklearn    << function = scorer_regression

<< track_header = True

<< metric = mae

<< greater_is_better = True

<< needs_threshold = False

<< needs_proba = False

<< kwargs = {}

>> id scorer

Note

The documentation page for function parameters: http://scikit-learn.org/0.15/modules/generated/sklearn.metrics.make_scorer.html#sklearn.metrics.make_scorer

corr

task:

Search

subtask:

evaluate

host:

pandas

function:

corr

input tokens (receivers):

df : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

output tokens (senders):

df : pandas dataframe

types: (“<class ‘pandas.core.frame.DataFrame’>”,)

required packages:

pandas, 0.20.3

config file view:

##

<< host = pandas    << function = corr

<< min_periods = 1

<< method = pearson

>> id df

>> id df

Note

The documentation page for function parameters: https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.corr.html