Potential issue in the merged contract timeseries

[kobi4000]

When examining contract price data for YENEUR#20241200.parquet, I've observed a potential issue with the merging of hourly and daily data.

pl.read_parquet('YENEUR#20241200.parquet').filter(pl.col('index').dt.date()==pl.date(2024,12,13))

OPEN	HIGH	LOW	FINAL	VOLUME	index
161.07	161.12	161.07	161.12	11.0	2024-12-13 14:30:00
161.08	161.1	161.08	161.08	5.0	2024-12-13 15:00:00
161.13	161.27	161.13	161.26	4.0	2024-12-13 16:00:00
161.37	161.37	161.37	161.37	3.0	2024-12-13 17:00:00
161.4	161.4	161.4	161.4	3.0	2024-12-13 19:00:00
161.07	161.4	161.07	161.28	26.0	2024-12-13 23:00:00

The last row appers to be daily aggregation and it seems keeping this in the merged timeseries is not too be helpful for downstream operations.

As my understanding EOD data is adjusted to 23:00 in syscore/dateutils.syscore/dateutils.py , and hourly and eod timesries is merged with syscore.pandas.frequency.merge_with_different_freq

Perhaps we could explore refining the merge logic to avoid including daily data when hourly data is available?