convert a column(string) in csv file to a tuple of ints

Question

Currently, I process a file using the csv module, which creates a list of dictionaries.

import csv
file = open('csvfile.csv')
lines = csv.reader(file)
header = next(lines)  # ['name', 'price', 'date']
# when I do the following
for line in lines:
    print(line)
# I get the following     
['xxxx', '5.00', '2/23/2023']

# assigning types to the columns to do type conversion using a function
types = [
    str,
    float,
    str  # this need to be a tuple
    # tried tuple(map(int, cannotchoosecolumn.split('/')))
    # did not work
]

# now to create a list of dicts
alist_of_dicts = [
    {
        name: func(val)
        for name, func, val in zip(header, types, line)
    }
    for line in lines
]

How would I select the third column str(2/23/2023) to change to a tuple(2, 21, 2007) using the format I am currently using?

Asked By: capnhud

||

Source

Answer 1

You can pass a function to your types list:

import datetime


def read_date(s):
    d = datetime.datetime.strptime(s, "%m/%d/%Y")
    return (d.month, d.day, d.year)


header = ["name", "price", "date"]
lines = [["xxxx", "5.00", "2/23/2023"]]
types = [
    str,
    float,
    read_date,
]

alist_of_dicts = [
    {name: func(val) for name, func, val in zip(header, types, line)} for line in lines
]

print(alist_of_dicts)
# prints: [{'name': 'xxxx', 'price': 5.0, 'date': (2, 23, 2023)}]

This is hard to understand code though. Instead, I recommend you use csv.DictReader to read the csv as a dictionary of strings -> strings, then transforming the columns

Answered By: Ben

Answer 2

Use a csv.DictReader and convert the columns as you read them:

import csv

with open('csvfile.csv', newline='') as file:
    a_list_of_dicts = []
    for line in csv.DictReader(file):
        line['price'] = float(line['price'])
        line['date'] = tuple(int(n) for n in line['date'].split('/'))
        a_list_of_dicts.append(line)

print(a_list_of_dicts)

csvfile.csv

name,price,date
xxxx,5.00,2/23/2023
yyyy,6.75,2/24/2023

Output:

[{'name': 'xxxx', 'price': 5.0, 'date': (2, 23, 2023)}, {'name': 'yyyy', 'price': 6.75, 'date': (2, 24, 2023)}]

Answered By: Mark Tolonen

convert a column(string) in csv file to a tuple of ints

Question:

Answers: