IMHO unacceptably large, but that's because Python doesn't know which parts of pandas it might need to execute the program.
A simpler alternative would be for Numpy and Pandas to provide its features as subpackages, like Airflow does: https://airflow.readthedocs.io/en/latest/installation.html