Queries from large csv files using dask are very slow

Question

I want to query only the rows for a specific user ID from a large csv file with 100 million rows that has multiple columns containing the user ID. There are about 400,000 user I