feyn.__future__.contrib.reference

function plot_kfold_traintest_score_results

def plot_kfold_traintest_score_results(
    rkf_results,
    train_lim=(0.0, 1.0),
    test_lim=(0.0, 1.0),
    figsize=(10, 5),
    title='Repeated k-fold results',
    ax=None,
    **kwargs_plot
)

Plots the results of train and test scores from the repeated k-fold
experiment run for any desired model(s).

Arguments:
    rkf_results {np.ndarray} -- array with k-fold results;
    must be shape (n_repeats, n_splits, 2)

Keyword Arguments:
    train_lim {tuple} -- plot range limit for training scores (default: {(0., 1.)})
    test_lim {tuple} -- plot range limit for testing scores (default: {(0., 1.)})
    figsize {tuple} -- size of the figure (default: {(10, 5)})
    title {str} -- title of the plot (default: {'Repeated k-fold results'})
    ax {matplotlib.axes._subplots.AxesSubplot} -- axes object (default: {None})

function plot_multiple_kfold_results

def plot_multiple_kfold_results(
    rkf_res_list,
    labels=['linear', 'tree', 'QLattice']
)

Function to plot multiple kfold analyses results.

Arguments:
    rkf_res_list {list} -- List of kfold results

Keyword Arguments:
    labels {list} -- List of labels corresponding to rkf_res_list
    (default: {['linear', 'tree', 'QLattice']})

function repeated_kfold_classification_analysis

def repeated_kfold_classification_analysis(
    data,
    target,
    model,
    n_splits=3,
    n_repeats=10,
    random_state=None,
    qlattice_updates=15,
    stypes={}
)

Repeates a k-fold cross-validation test for either a
Linear model, a Decision Tree or the QLattice. Idea of usage
is to check whether data has any signal.

Arguments:
    data {DataFrame} -- dataset to be analyzed
    target {str} -- output variable of dataset
    model {str} -- type of model; can be 'linear', 'tree' or feyn.QLattice object

Keyword Arguments:
    n_splits {int} -- number of splits in a single k-fold round (default: {3})
    n_repeats {int} -- total number of k-fold rounds (default: {10})
    random_state {int} -- random state of the repeated k-fold split (default: {None})
    qlattice_updates {int} -- number of QLattice updates (default: {15})
    stypes {dict} -- semantical types for the QLattice (default: {dict()})

function repeated_kfold_regression_analysis

def repeated_kfold_regression_analysis(
    data,
    target,
    model,
    n_splits=3,
    n_repeats=10,
    random_state=None,
    qlattice_updates=15,
    stypes={}
)

Repeates a k-fold cross-validation test for either a
Linear model, a Decision Tree or the QLattice. Idea of usage
is to check whether data has any signal.

Arguments:
    data {DataFrame} -- dataset to be analyzed
    target {str} -- output variable of dataset
    model {str} -- type of model; can be 'linear', 'tree' or feyn.QLattice object

Keyword Arguments:
    n_splits {int} -- number of splits in a single k-fold round (default: {3})
    n_repeats {int} -- total number of k-fold rounds (default: {10})
    random_state {int} -- random state of the repeated k-fold split (default: {None})
    qlattice_updates {int} -- number of QLattice updates (default: {15})
    stypes {dict} -- semantical types for the QLattice (default: {dict()})