How to use the rsmtool.test_utils.do_run_experiment function in rsmtool

def test_run_experiment_duplicate_feature_names():

    # rsmtool experiment with duplicate feature names
    source = 'lr-with-duplicate-feature-names'
    experiment_id = 'lr_with_duplicate_feature_names'
    config_file = join(rsmtool_test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))
    do_run_experiment(source, experiment_id, config_file)

def test_run_experiment_wrong_train_file_path():
    # basic experiment with the path in train_file field pointing to
    # a non-existing file
    source = 'lr-wrong-path'
    experiment_id = 'lr'
    config_file = join(rsmtool_test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))
    do_run_experiment(source, experiment_id, config_file)

def test_run_experiment_adaboost():

    # basic experiment with AdaBoostRegressor model

    source = 'adaboost'
    experiment_id = 'AdaBoost'
    config_file = join(test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))
    do_run_experiment(source, experiment_id, config_file)

    output_dir = join('test_outputs', source, 'output')
    expected_output_dir = join(test_dir, 'data', 'experiments', source, 'output')
    html_report = join('test_outputs', source, 'report', '{}_report.html'.format(experiment_id))

    csv_files = glob(join(output_dir, '*.csv'))
    for csv_file in csv_files:
        csv_filename = basename(csv_file)
        expected_csv_file = join(expected_output_dir, csv_filename)

        if exists(expected_csv_file):
            yield check_csv_output, csv_file, expected_csv_file

    yield check_all_csv_exist, csv_files, experiment_id, 'skll'
    yield check_report, html_report

def test_run_experiment_lr_subgroups_with_h2():

    # basic experiment with subgroups and second
    # rater analyses

    source = 'lr-subgroups-with-h2'
    experiment_id = 'lr_subgroups_with_h2'
    config_file = join(test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))
    do_run_experiment(source, experiment_id, config_file)

    output_dir = join('test_outputs', source, 'output')
    expected_output_dir = join(test_dir, 'data', 'experiments', source, 'output')
    html_report = join('test_outputs', source, 'report', '{}_report.html'.format(experiment_id))

    csv_files = glob(join(output_dir, '*.csv'))
    for csv_file in csv_files:
        csv_filename = basename(csv_file)
        expected_csv_file = join(expected_output_dir, csv_filename)

        if exists(expected_csv_file):
            yield check_csv_output, csv_file, expected_csv_file

    yield check_all_csv_exist, csv_files, experiment_id, 'rsmtool'
    yield check_consistency_files_exist, csv_files, experiment_id
    yield check_scaled_coefficients, source, experiment_id

def test_run_experiment_lr_with_sc2_as_feature_name():

    # rsmtool experiment with sc2 used as a feature name
    # when the user also requests h2 analysis using a different
    # column
    source = 'lr-with-sc2-as-feature-name'
    experiment_id = 'lr_with_sc2_as_feature_name'
    config_file = join(test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))
    do_run_experiment(source, experiment_id, config_file)

def test_run_experiment_lr_with_large_integer_value():
    # basic experiment with LinearRegression model with some
    # values for FEATURE7 (the integer feature) being
    # outside +/-4 SD.

    source = 'lr-with-large-integer-value'
    experiment_id = 'lr_with_large_integer_value'
    config_file = join(test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))
    do_run_experiment(source, experiment_id, config_file)
    output_dir = join('test_outputs', source, 'output')
    expected_output_dir = join(test_dir, 'data', 'experiments', source, 'output')
    html_report = join('test_outputs', source, 'report', '{}_report.html'.format(experiment_id))

    csv_files = glob(join(output_dir, '*.csv'))
    for csv_file in csv_files:
        csv_filename = basename(csv_file)
        expected_csv_file = join(expected_output_dir, csv_filename)

        if exists(expected_csv_file):
            yield check_csv_output, csv_file, expected_csv_file

    yield check_all_csv_exist, csv_files, experiment_id, 'rsmtool'
    yield check_scaled_coefficients, source, experiment_id
    yield check_report, html_report

def test_run_experiment_lr_feature_json():
    # basic experiment with a LinearRegression model but using
    # feature json file

    source = 'lr-feature-json'
    experiment_id = 'lr'
    config_file = join(rsmtool_test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))

    # run this experiment but suppress the expected deprecation warnings
    do_run_experiment(source,
                      experiment_id,
                      config_file,
                      suppress_warnings_for=[DeprecationWarning])

def test_run_experiment_lr_tsv_input_and_subset_files():

    # rsmtool experiment with input files in .tsv format
    # including a feature subset file in .tsv format

    source = 'lr-tsv-input-and-subset-files'
    experiment_id = 'lr_tsv_input_and_subset_files'
    config_file = join(test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))
    do_run_experiment(source, experiment_id, config_file)

    output_dir = join('test_outputs', source, 'output')
    expected_output_dir = join(test_dir, 'data', 'experiments', source, 'output')
    html_report = join('test_outputs', source, 'report', '{}_report.html'.format(experiment_id))

    csv_files = glob(join(output_dir, '*.csv'))
    for csv_file in csv_files:
        csv_filename = basename(csv_file)
        expected_csv_file = join(expected_output_dir, csv_filename)

        if exists(expected_csv_file):
            yield check_csv_output, csv_file, expected_csv_file

    yield check_all_csv_exist, csv_files, experiment_id, 'rsmtool'
    yield check_scaled_coefficients, source, experiment_id
    yield check_report, html_report

def test_run_experiment_svr():

    # basic experiment with svr model

    source = 'svr'
    experiment_id = 'SVR'
    config_file = join(test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))
    do_run_experiment(source, experiment_id, config_file)

    output_dir = join('test_outputs', source, 'output')
    expected_output_dir = join(test_dir, 'data', 'experiments', source, 'output')
    html_report = join('test_outputs', source, 'report', '{}_report.html'.format(experiment_id))

    csv_files = glob(join(output_dir, '*.csv'))
    for csv_file in csv_files:
        csv_filename = basename(csv_file)
        expected_csv_file = join(expected_output_dir, csv_filename)

        if exists(expected_csv_file):
            yield check_csv_output, csv_file, expected_csv_file

    yield check_all_csv_exist, csv_files, experiment_id, 'skll'
    yield check_report, html_report

def test_run_experiment_lr():
    # basic experiment with a LinearRegression model

    source = 'lr'
    experiment_id = 'lr'
    config_file = join(test_dir,
                       'data',
                       'experiments',
                       source,
                       '{}.json'.format(experiment_id))
    do_run_experiment(source, experiment_id, config_file)
    output_dir = join('test_outputs', source, 'output')
    expected_output_dir = join(test_dir, 'data', 'experiments', source, 'output')
    html_report = join('test_outputs', source, 'report', '{}_report.html'.format(experiment_id))

    csv_files = glob(join(output_dir, '*.csv'))
    for csv_file in csv_files:
        csv_filename = basename(csv_file)
        expected_csv_file = join(expected_output_dir, csv_filename)

        if exists(expected_csv_file):
            yield check_csv_output, csv_file, expected_csv_file

    yield check_all_csv_exist, csv_files, experiment_id, 'rsmtool'
    yield check_scaled_coefficients, source, experiment_id
    yield check_report, html_report