Pandas rolling max for time series data

Question

I get 2 different behaviours when applying rolling("1D").max() on 2 datasets in Jupyter notebook.

I need to calculate rolling max for each day.

S