How do I drop nan, inf, and -inf values from a DataFrame without resetting mode.use_inf_as_null?
Can I tell dropna to include inf in its definition of missing values so that the following works?
df.dropna(subset=["col1", "col2"], how="all")