def test_overall_accuracy_Net4(): pathToPred = ds.datasets['Net4'].get_outFile() + \ '_priorPerct100_weight0.01__scaled_FeatureScaling::InfOnly' AUROC, AUPR, fpr, tpr = get_accuracy_realData(network='Net4', predFile=pathToPred, visualize=False) assert_almost_equal(AUROC, 0.31) assert_almost_equal(AUPR, 0.75)
if __name__ == "__main__": datasetName = 'Root' #'Net1', 'Net3_conn_final', 'Net4', 'Grene' alphas_small = [0.001, 0.01, 0.05, 0.1] alphas_large = [0.1, 0.2, 0.4, 0.6, 0.8, 1, 2, 3] # params = OrderedDict([('fileName', ''), ('priorPercent', 0), ('falsePriorRatio', 0), ('priorWeight', 0.01), ('alpha', 0.1), ('l1_ratio', 0.5), ('isCV', 1), ('fit_intercept', 1), ('scaleX', 1), ('halfLife', 10), ('method', 'FeatureScaling'), ('freeCV', 0)]) # params = OrderedDict([('fileName', ''), ('priorPercent', 100), ('falsePriorRatio', 0), # ('priorWeight', 0.01), # ('alpha', 0.1), ('l1_ratio', 0.5), ('isCV', 1), ('fit_intercept', 1), # ('scaleX', 1), ('halfLife', 10), ('method', 'PenaltyScaling'), ('freeCV', 1)]) filename = run_Peak_test(datasetName, params, alphas_small) pathToPred = ds.datasets['Root'].get_outFile() + filename AUROC, AUPR, _, _ = get_accuracy_realData(network=datasetName, predFile=pathToPred, visualize=False) pathToCombined = pathToPred[:-9] + "::Combined" AUROC, AUPR, _, _ = get_accuracy_realData(network=datasetName, predFile=pathToCombined, visualize=False) mapGeneNames(datasetName='Root', filename=pathToPred) # print(paramsFileName) # filename = ds.datasets['Root'].get_outFile() + '_priorPerct100_weight0.01__scaled__Nov-18_18-22--03FeatureScaling::InfOnly' # mapGeneNames(datasetName='Root', filename=filename)
params = OrderedDict([('fileName', ''), ('priorPercent', 100), ('falsePriorRatio', 0), ('priorWeight', 0.01), ('priorFile', priorFile), ('pkEachGene', 1), ('alpha', 0.1), ('l1_ratio', 0.5), ('isCV', 1), ('fit_intercept', 1), ('scaleX', 1), ('halfLife', 10), ('method', 'PenaltyScaling'), ('freeCV', 0)]) # params = OrderedDict([('fileName', ''), ('priorPercent', 100), ('falsePriorRatio', 0), # ('priorWeight', 0.01), # ('alpha', 0.1), ('l1_ratio', 0.5), ('isCV', 1), ('fit_intercept', 1), # ('scaleX', 1), ('halfLife', 10), ('method', 'PenaltyScaling'), ('freeCV', 1)]) filename = run_Peak_test(datasetName, params, alphas_small) pathToPred = ds.datasets[datasetName].get_outFile() + filename AUROC, AUPR, fpr, tpr = get_accuracy_realData(network=datasetName, predFile=pathToPred, visualize=False) print('fpr', fpr) print('tpr', tpr) pathToCombined = pathToPred[:-9] + "::Combined" AUROC, AUPR, _, _ = get_accuracy_realData(network=datasetName, predFile=pathToCombined, visualize=False) # mapGeneNames(datasetName='Root', filename=pathToPred) # print(paramsFileName) # filename = ds.datasets['Root'].get_outFile() + '_priorPerct100_weight0.01__scaled__Nov-18_18-22--03FeatureScaling::InfOnly' # mapGeneNames(datasetName='Root', filename=filename)