Today, you will learn how to read and write CSV files in Python. CSV (comma-separated values) are one of the most widely used types to store table data. To read and write CSV files, we will be using the csv module from Python.

Reading a CSV

If you simply want to open a CSV and print out the contents in Python, you can do so by running the following code:

import csv

with open('YOUR_CSV.csv', 'r') as file:
    reader = csv.reader(file)
    for row in reader:
        print(row)

Writing a CSV

Using arrays

If you want to write a CSV using arrays:

• Import the csv module.
• Define an array for the header.
• Define an array of arrays for the rows.
• Create a new CSV file in Python using csv.writer.
• Write the header with writerow.
• Write the rows using writerows.

import csv 

header = ['PAIR', 'PRICE'] 

rows = [ ['USDCAD', 1.27], 
         ['GBPUSD', 1.38], 
         ['EURUSD', 1.19]] 
  
with open("forex_prices.csv", 'w', newline='') as file: 
    csvwriter = csv.writer(file) 
    csvwriter.writerow(header) 
    csvwriter.writerows(rows)

Using dictionaries

Saving an array of dictionaries to a CSV file is similar to saving an array of rows:

• Import the csv module.
• Define the header of your table using an array.
• Define the array of dictionaries (each row).
• Create a new CSV file using csv.DictWriter passing in the file and set fieldnames = header.
• Call writeheader() to write the header.
• Call writerows and pass the array of dictionaries.

import csv 

header = ['PAIR', 'PRICE'] 
		 
rows_dict =[{'PAIR': 'USDCAD', 'PRICE': 1.27}, 
			{'PAIR': 'GBPUSD', 'PRICE': 1.38},
			{'PAIR': 'EURUSD', 'PRICE': 1.19}] 
  
with open("forex_prices.csv", 'w', newline='') as file: 
	csvwriter = csv.DictWriter(file, fieldnames = header) 
	csvwriter.writeheader() 
	csvwriter.writerows(rows_dict) 

Quoting

When reading and writing a CSV, you can specify what type of quoting you want. This means that if you want to put quotes around all data, or only certain data you can do this using quoting. There are 4 types of quoting constants in the CSV module which can be passed to csv.writer and csv.DictWriter:

QUOTE_NONE – Quotes nothing.
QUOTE_NONNUMERIC – Quotes anything that is not an int or float
QUOTE_MINIMAL – Quotes every field that contains a special character
QUOTE_ALL – Quotes every field

Example

If you want to quote every field, you would create a new csv.writer as follows:

import csv 

header = ['PAIR', 'PRICE'] 

rows = [ ['USDCAD', 1.27], 
         ['GBPUSD', 1.38], 
         ['EURUSD', 1.19]] 
  
with open("forex_prices.csv", 'w', newline='') as file: 
    csvwriter = csv.writer(file, quoting=csv.QUOTE_ALL) 
    csvwriter.writerow(header) 
    csvwriter.writerows(rows)

Output:

"PAIR","PRICE"
"USDCAD","1.27"
"GBPUSD","1.38"
"EURUSD","1.19"

As you can see above, everything has been quoted. Let’s see how this differs from using QUOTE_NONNUMERIC:

import csv 

header = ['PAIR', 'PRICE'] 

rows = [ ['USDCAD', 1.27], 
         ['GBPUSD', 1.38], 
         ['EURUSD', 1.19]] 
  
with open("forex_prices.csv", 'w', newline='') as file: 
    csvwriter = csv.writer(file, quoting=csv.QUOTE_NONNUMERIC) 
    csvwriter.writerow(header) 
    csvwriter.writerows(rows)

Output:

"PAIR","PRICE"
"USDCAD",1.27
"GBPUSD",1.38
"EURUSD",1.19

As you can see, Only the PAIR field has been quoted as it is not an int or float.

Dialect

The csv module offers a concept called dialects. A dialect allows you to modify the format of reading/writing a CSV file. For instance, if I wanted to use ‘;’ instead of ‘,’ as my delimiter, I could create a dialect for this.

Start by defining your dialect name, specifying the delimiter parameter as ‘;’ and registering it:

csv.register_dialect('semi-colon', delimiter=';')

When we call csv.writer or csv.reader you can specify the dialect by using the dialect name you defined above. In my case this is ‘semi-colon’:

csvwriter = csv.writer(file, dialect='semi-colon') 

Example

import csv 

header = ['PAIR', 'PRICE'] 

rows = [ ['USDCAD', 1.27], 
         ['GBPUSD', 1.38], 
         ['EURUSD', 1.19]] 

csv.register_dialect('semi-colon', delimiter=';')
	
with open("forex_prices.csv", 'w', newline='') as file: 
    csvwriter = csv.writer(file, dialect='semi-colon') 
    csvwriter.writerow(header) 
    csvwriter.writerows(rows)

Now when this code is run, instead of the fields being separated by ‘,’ – they will be separated by ‘;’

PAIR;PRICE
USDCAD;1.27
GBPUSD;1.38
EURUSD;1.19

Dialect Parameters

A comprehensive list of every parameter you can use in a Dialect and their default values:

That’s all for how to read and write CSV files in Python! As always, if you have any questions or comments please feel free to post them below. Additionally, if you run into any issues please let me know.

If you’re interested in learning Python I highly recommend this book. In the first half of the book, you”ll learn basic programming concepts, such as variables, lists, classes, and loops, and practice writing clean code with exercises for each topic. In the second half, you”ll put your new knowledge into practice with three substantial projects: a Space Invaders-inspired arcade game, a set of data visualizations with Python”s handy libraries, and a simple web app you can deploy online. Get it here.

Check out these other Python tutorials!