Reading NC Voter data from TXT with minimal memory

ncdata.py

import pandas as pd

txtfile = 'ncvoter_Statewide.txt'
coltypes = {
    'county_id': 'int',
    'voter_reg_num': 'int',
    'birth_year': 'int',
    'age_at_year_end': 'int',
    'county_desc': 'category',
    'voter_status_desc': 'category',
    'voter_status_reason_desc': 'category',
    'res_city_desc': 'category',
    'precinct_abbrv': 'category',
    'precinct_desc': 'category',
    'municipality_abbrv': 'category',
    'municipality_desc': 'category',
    'ward_abbrv': 'category',
    'ward_desc': 'category',
    'cong_dist_abbrv': 'category',
    'super_court_abbrv': 'category',
    'judic_dist_abbrv': 'category',
    'nc_senate_abbrv': 'category',
    'nc_house_abbrv': 'category',
    'county_commiss_abbrv': 'category',
    'county_commiss_desc': 'category',
    'township_abbrv': 'category',
    'township_desc': 'category',
    'school_dist_abbrv': 'category',
    'school_dist_desc': 'category',
    'fire_dist_abbrv': 'category',
    'fire_dist_desc': 'category',
    'water_dist_abbrv': 'category',
    'water_dist_desc': 'category',
    'sewer_dist_abbrv': 'category',
    'sewer_dist_desc': 'category',
    'sanit_dist_abbrv': 'category',
    'sanit_dist_desc': 'category',
    'rescue_dist_abbrv': 'category',
    'rescue_dist_desc': 'category',
    'munic_dist_abbrv': 'category',
    'munic_dist_desc': 'category',
    'dist_1_abbrv': 'category',
    'dist_1_desc': 'category',
    'vtd_abbrv': 'category',
    'vtd_desc': 'category',
}

df = pd.read_csv(txtfile, sep='\t', encoding='ISO-8859-1', dtype=coltypes)
df.head()