MultiAssetPurgedKFold and MultiAssetCombinatorialPurgedKFold are throwing TypeError: '>=' not supported between instances of 'tuple' and 'str'

[brandonyano]

Describe the bug
MultiAssetPurgedKFold and MultiAssetCombinatorialPurgedKFold are throwing TypeError: '>=' not supported between instances of 'tuple' and 'str' when using a stacked multi-asset dataframe, the error doesn't occur if you only send one asset at a time.

To Reproduce

idx = pd.MultiIndex.from_product([['AAPL', 'MSFT', 'TSLA'],
                                  ['2006-03-28 10:46:00', '2006-03-28 14:35:00', '2006-03-28 21:08:00']],
                                 names=['asset', 'date_time'])

t1_events = pd.DataFrame('2006-03-29 10:46:00', idx, ['t1'])
prices = pd.DataFrame(100, idx, ['price'])

cv = MultiAssetPurgedKFold(n_splits=5, samples_info_sets=t1_events["t1"], pct_embargo=0.01)

splits = []
for train, test in cv.split(prices):
    splits.append((train, test))

Screenshots

Desktop (please complete the following information):

• mlfinlab v1.6.0