How do you check if multiple timeseries settings are all-NaN in a Dataframe? (fastest way)

Question

i have a series of timeindices Ts and a dataframe of multiple columns df.

Ts.index are the startidices of a startwindow and