CSV

from datetime import datetime

dates = {}
with open('v.csv') as f:  # Path to csv file
    for line in f.readlines():
        values = line.strip().split(',')  # List of comma-separated values, strip() used to get rid of '\n' symbol in the end
        user_id = values[3]  # Index of id
        date = datetime.strptime(values[6], '"%Y-%m-%d %H:%M:%S"')  # Parsing sate and time from a string to datetime object
        if user_id in dates:  # If we already have a date for this user
            dates[user_id] = min(date, dates[user_id])  # Minimum of date we just read and the one we recorded before
        else:
            dates[user_id] = date  # Date we just read

for item in dates.items():
    print("User with id {} made first purchase on {}".format(item[0], item[1]))