get_ncv_table.py

def ncv_runs_table(number_of_folds):
    '''
    ez function to generate nested-cross-validation runs
    mostly for explanation purposes
    '''
    
    from itertools import combinations

    folds = range(number_of_folds)
    lst = []
    c = 0
    for i in list(combinations(folds, 2)):
        d = {}
        d['validation'] = [i[0]]
        d['test'] = [i[1]]
        d['train'] = list(set(folds).difference(set(i)))
        c = c + 1
        d['run'] = f'run{c}'
        lst.append(d)

        d2 = {}
        d2['validation'] = [i[1]]
        d2['test'] = [i[0]]
        d2['train'] = list(set(folds).difference(set(i)))
        c = c + 1
        d2['run'] = f'run{c}'
        lst.append(d2)
    return pd.DataFrame(lst)