dict(model_solver=al_asgl_model, weight_calculator=al_asgl_wc_pls_1),
dict(model_solver=asgl_model, weight_calculator=asgl_wc_pls_1),
dict(model_solver=al_asgl_model, weight_calculator=al_asgl_wc_pls_d),
dict(model_solver=asgl_model, weight_calculator=asgl_wc_pls_d)
]
results = gf.automatic_simulator(data_param=data_param,
model_selection_param=model_selection_param,
model_param=model_param,
data_generator=data_generator,
n_repetitions=n_repetitions,
folder=folder)
########################################################################################################################
results = af.simulation_results_to_tables(results=results,
from_file=False,
table_format='row_models')
sorting = [
'lasso', 'sgl', 'al_asgl_pca_d', 'al_asgl_pca_1', 'al_asgl_pls_d',
'al_asgl_pls_1', 'asgl_pca_d', 'asgl_pca_1', 'asgl_pls_d', 'asgl_pls_1'
]
gc.boxplot_creator_by_metric(
results=results,
interesting_metrics=['final_error', 'non_zero_pred_beta'],
figsize=(25, 10),
sorting=sorting)
########################################################################################################################
dict(model_solver=asgl_model, weight_calculator=asgl_wc_pls_d)
]
results7 = gf.automatic_analyzer(dataset=dataset,
model_selection_param=model_selection_param,
model_param=model_param,
n_repetitions=n_repetitions,
folder=folder)
########################################################################################################################
"""
Processing the results obtained
"""
gc.boxplot_creator_by_metric(
results=results,
interesting_metrics=['final_error', 'non_zero_pred_beta'],
figsize=(25, 10),
sorting=['lasso', 'sgl', 'asgl_pca_d', 'asgl_pls_d'])
significance = gc.variables_probability_heatmap(
results=results, sorting=['asgl_pca_d', 'asgl_pls_d', 'sgl', 'lasso'])
model_names = ['lasso', 'sgl', 'asgl_pca_d', 'asgl_pls_d']
probability_of_significance = significance['probability_of_significance']
# Number of genes above a threshold
threshold = 0.5
for i in range(len(model_names)):
tmp_prob = probability_of_significance[i, :]
num_genes = len(np.where(tmp_prob >= threshold)[0])
print("Model: {}. Threshold: {}. Number of genes: {}".format(
model_names[i], threshold, num_genes))