def test_process_emodel_no_exemplars():
# input parameters
emodel = 'emodel1'
mtypes = ['mtype1', 'mtype2']
scores_dict = {'emodel': [emodel, emodel],
'is_exemplar': [0, 0], 'is_repaired': [1, 0],
'is_original': [1, 0],
'opt_scores': [json.dumps({'Step1.SpikeCount': 2.0}),
json.dumps({'Step1.SpikeCount': 2.0})],
'scores': [json.dumps({'Step1.SpikeCount': 2.0}),
json.dumps({'Step1.SpikeCount': 2.0})],
'etype': ['etype1', 'etype2'],
'fullmtype': mtypes,
'mtype': mtypes,
'extra_values': [json.dumps({'threshold_current': 0.0,
'holding_current': 0.0}),
json.dumps({'threshold_current': 0.0,
'holding_current': 0.0})],
'morph_name': ['morph1', 'morph1'],
'layer': ['layer_1', 'layer_1']
}
scores = pandas.DataFrame(scores_dict)
score_values_dict = {'Step1.SpikeCount': [2.0, 2.0]}
score_values = pandas.DataFrame(score_values_dict)
to_skip_patterns = []
regex_all = re.compile('.*')
megate_patterns = [{'megate_feature_threshold': {'megate_threshold': 5,
'features': regex_all},
'emodel': regex_all, 'fullmtype': regex_all,
'etype': regex_all}]
skip_repaired_exemplar = False
enable_check_opt_scores = True
select_best_perc = None
# run function
ret = table_processing.process_emodel((
emodel, scores, score_values, to_skip_patterns, megate_patterns,
skip_repaired_exemplar, enable_check_opt_scores, select_best_perc))
# verify results
nt.assert_is_none(ret)
def test_process_emodel():
# input parameters
emodel = 'emodel1'
mtypes = ['mtype1', 'mtype2']
scores_dict = {'emodel': [emodel, emodel],
'is_exemplar': [1, 0], 'is_repaired': [1, 0],
'is_original': [0, 0],
'opt_scores': [json.dumps({'Step1.SpikeCount': 2.0}),
json.dumps({'Step1.SpikeCount': 2.0})],
'scores': [json.dumps({'Step1.SpikeCount': 2.0}),
json.dumps({'Step1.SpikeCount': 2.0})],
'etype': ['etype1', 'etype2'],
'fullmtype': mtypes,
'mtype': mtypes,
'extra_values': [json.dumps({'threshold_current': 0.0,
'holding_current': 0.0}),
json.dumps({'threshold_current': 0.0,
'holding_current': 0.0})],
'morph_name': ['morph1', 'morph1'],
'layer': ['layer_1', 'layer_1']
}
scores = pandas.DataFrame(scores_dict)
score_values_dict = {'Step1.SpikeCount': [2.0, 2.0]}
score_values = pandas.DataFrame(score_values_dict)
to_skip_patterns = []
regex_all = re.compile('.*')
megate_patterns = [{'megate_feature_threshold': {'megate_threshold': 5,
'features': regex_all},
'emodel': regex_all, 'fullmtype': regex_all,
'etype': regex_all}]
skip_repaired_exemplar = False
enable_check_opt_scores = True
select_best_perc = None
emodel, emodel_info = table_processing.process_emodel(
(emodel, scores, score_values, to_skip_patterns, megate_patterns,
skip_repaired_exemplar, enable_check_opt_scores, select_best_perc))
(emodel_ext_neurondb, emodel_megate_pass, emodel_score_values,
mtypes, emodel_megate_passed_all, emodel_median_scores,
passed_combos) = emodel_info
nt.assert_equal(emodel, emodel)
# expected results
columns = ['morph_name', 'layer', 'fullmtype', 'etype', 'emodel',
'combo_name', 'threshold_current', 'holding_current']
db_dict = {'morph_name': 'morph1',
'layer': 'layer_1',
'fullmtype': 'mtype2',
'etype': 'etype2',
'emodel': 'emodel1',
'threshold_current': 0.0,
'holding_current': 0.0,
'combo_name': 'emodel1_mtype2_layer_1_morph1'
}
exp_db = pandas.DataFrame(data=db_dict, columns=columns, index=[1])
columns = ['Step1.SpikeCount', 'Passed all']
megate_scores_dict = {'Step1.SpikeCount': True, 'Passed all': True}
exp_megate_scores = pandas.DataFrame(data=megate_scores_dict,
columns=columns, index=[1])
exp_score_values = pandas.DataFrame({'Step1.SpikeCount': 2.0}, index=[1])
exp_mtypes = pandas.Series('mtype2', index=[1], name='fullmtype')
# verify results
pandas.util.testing.assert_frame_equal(emodel_ext_neurondb, exp_db)
pandas.util.testing.assert_frame_equal(
emodel_megate_pass, exp_megate_scores)
pandas.util.testing.assert_frame_equal(
emodel_score_values, exp_score_values)
pandas.util.testing.assert_series_equal(mtypes, exp_mtypes)