isna() (alias isnull()) returns a boolean mask indicating missing values (NaN, None, NaT). Combined with sum() it counts missing values per column, and with mean() it gives their proportion. notna() is its inverse.
Example
df.isna().sum()