How to use the pydash.difference function in pydash

def test_difference(case, expected):
    assert _.difference(*case) == expected

def calc_experiment_df(trial_data_dict, info_prepath=None):
    '''Collect all trial data (metrics and config) from trials into a dataframe'''
    experiment_df = pd.DataFrame(trial_data_dict).transpose()
    cols = METRICS_COLS
    config_cols = sorted(ps.difference(experiment_df.columns.tolist(), cols))
    sorted_cols = config_cols + cols
    experiment_df = experiment_df.reindex(sorted_cols, axis=1)
    experiment_df.sort_values(by=['strength'], ascending=False, inplace=True)
    if info_prepath is not None:
        util.write(experiment_df, f'{info_prepath}_experiment_df.csv')
        # save important metrics in info_prepath directly
        util.write(experiment_df, f'{info_prepath.replace("info/", "")}_experiment_df.csv')
    return experiment_df

def plot_experiment(experiment_spec, experiment_df, metrics_cols):
    '''
    Plot the metrics vs. specs parameters of an experiment, where each point is a trial.
    ref colors: https://plot.ly/python/heatmaps-contours-and-2dhistograms-tutorial/#plotlys-predefined-color-scales
    '''
    y_cols = metrics_cols
    x_cols = ps.difference(experiment_df.columns.tolist(), y_cols + ['trial'])
    fig = tools.make_subplots(rows=len(y_cols), cols=len(x_cols), shared_xaxes=True, shared_yaxes=True, print_grid=False)
    strength_sr = experiment_df['strength']
    min_strength, max_strength = strength_sr.min(), strength_sr.max()
    for row_idx, y in enumerate(y_cols):
        for col_idx, x in enumerate(x_cols):
            x_sr = experiment_df[x]
            guard_cat_x = x_sr.astype(str) if x_sr.dtype == 'object' else x_sr
            trace = go.Scatter(
                y=experiment_df[y], yaxis=f'y{row_idx+1}',
                x=guard_cat_x, xaxis=f'x{col_idx+1}',
                showlegend=False, mode='markers',
                marker={
                    'symbol': 'circle-open-dot', 'color': strength_sr, 'opacity': 0.5,
                    # dump first portion of colorscale that is too bright
                    'cmin': min_strength - 0.5 * (max_strength - min_strength), 'cmax': max_strength,
                    'colorscale': 'YlGnBu', 'reversescale': True

def retro_analyze_trials(predir):
    '''Retro analyze all trials'''
    logger.info('Running retro_analyze_trials')
    session_spec_paths = glob(f'{predir}/*_s*_spec.json')
    # remove session spec paths
    trial_spec_paths = ps.difference(glob(f'{predir}/*_t*_spec.json'), session_spec_paths)
    util.parallelize(_retro_analyze_trial, [(p,) for p in trial_spec_paths], num_cpus=util.NUM_CPUS)

def retro_analyze_experiment(predir):
    '''Retro analyze an experiment'''
    logger.info('Running retro_analyze_experiment')
    trial_spec_paths = glob(f'{predir}/*_t*_spec.json')
    # remove trial and session spec paths
    experiment_spec_paths = ps.difference(glob(f'{predir}/*_spec.json'), trial_spec_paths)
    experiment_spec_path = experiment_spec_paths[0]
    spec = util.read(experiment_spec_path)
    info_prepath = spec['meta']['info_prepath']
    if os.path.exists(f'{info_prepath}_trial_data_dict.json'):
        return  # only run analysis if experiment had been ran
    trial_data_dict = util.read(f'{info_prepath}_trial_data_dict.json')
    analysis.analyze_experiment(spec, trial_data_dict)

def plot_experiment(experiment_spec, experiment_df, metrics_cols):
    '''
    Plot the metrics vs. specs parameters of an experiment, where each point is a trial.
    ref colors: https://plot.ly/python/heatmaps-contours-and-2dhistograms-tutorial/#plotlys-predefined-color-scales
    '''
    y_cols = metrics_cols
    x_cols = ps.difference(experiment_df.columns.tolist(), y_cols + ['trial'])
    fig = tools.make_subplots(rows=len(y_cols), cols=len(x_cols), shared_xaxes=True, shared_yaxes=True, print_grid=False)
    strength_sr = experiment_df['strength']
    min_strength, max_strength = strength_sr.min(), strength_sr.max()
    for row_idx, y in enumerate(y_cols):
        for col_idx, x in enumerate(x_cols):
            x_sr = experiment_df[x]
            guard_cat_x = x_sr.astype(str) if x_sr.dtype == 'object' else x_sr
            trace = go.Scatter(
                y=experiment_df[y], yaxis=f'y{row_idx+1}',
                x=guard_cat_x, xaxis=f'x{col_idx+1}',
                showlegend=False, mode='markers',
                marker={
                    'symbol': 'circle-open-dot', 'color': strength_sr, 'opacity': 0.5,
                    # dump first portion of colorscale that is too bright
                    'cmin': min_strength - 0.5 * (max_strength - min_strength), 'cmax': max_strength,
                    'colorscale': 'YlGnBu', 'reversescale': True