Python Pandas - Read CSV

---

---

Reading data from a CSV (Comma-Separated Values) file is one of the most common tasks in data analysis and data science. Python's Pandas library provides a flexible read_csv() method for reading data from CSV files into a DataFrame.

This method simplifies the process of loading the data from a CSV file into a DataFrame (a 2D table-like structure with labeled rows and columns).

In this tutorial, we will learn various aspects of reading CSV files with Pandas, including advanced features like controlling the number of rows to read, parsing dates, handling missing data, and more.

Introduction to read_csv() Method

The pandas.read_csv() method in the Pandas library used to read a CSV file and converts the data into a Pandas DataFrame object. It can read files stored locally or remotely using supported URL schemes like HTTP, FTP, or cloud storage paths (e.g., S3).

This method supports a wide range of parameters for customizing headers, delimiters, and data types.

Key Features of read_csv() Method −

• This method can handle large datasets efficiently.

• It supports different delimiters and is not limited to reading comma-separated values.

• It supports reading data from different sources like, URLs, local files, and file-like objects.

• This method handles missing values effectively.

Syntax

Below is the syntax of the read_csv() method −

pandas.read_csv(filepath_or_buffer, sep=',', header='infer', names=None, index_col=None, usecols=None, dtype=None, ...)

Here are some commonly used parameters of this method −

• filepath_or_buffer: The file path, URL, or file-like object to read CSV file from.

• sep or delimiter: Specifies the character or regex pattern used to separate values in the file. Default is , (comma).

• header: Specifies the row number to use as column names.

• names: A list of column names to use when there is no header row.

Example

Here is a basic example of using the pandas.read_csv() method for reading a CSV data from local file.

import pandas as pd

# Read a CSV file into a DataFrame
data = pd.read_csv('data.csv')

# Display the first few rows
print(data.head())

Following is the output of the above code −

	Name	Salary
0	Ravi	50000
1	Priya	45000
2	Kiran	65000
3	NaN	55000

Reading Large CSV Files in Chunks

For reading large CSV files in small blocks, you can use the chunksize parameter of the read_csv() method. This parameter reads data into smaller, more manageable blocks of given size.

Example

This example shows loading the CSV data into the Pandas DataFrame of smaller chunks using the chunksize parameter.

import pandas as pd

url ="https://raw.usercontent.com/Opensourcefordatascience/Data-sets/master/blood_pressure.csv"

# Read the CSV data in chunks
chunks = pd.read_csv(url, chunksize=50)

for chunk in chunks:
   print("Output DataFrame:")
   print(chunk.head())

When we run above program, it will produces following result −

Output DataFrame:

	patient	sex	agegrp	bp_before	bp_after
0	1	Male	30-45	143	153
1	2	Male	30-45	163	170
2	3	Male	30-45	153	168
3	4	Male	30-45	153	142
4	5	Male	30-45	146	141

Output DataFrame:

	patient	sex	agegrp	bp_before	bp_after
50	51	Male	60+	175	146
51	52	Male	60+	175	160
52	53	Male	60+	172	175
53	54	Male	60+	173	163
54	55	Male	60+	170	185

Output DataFrame:

	patient	sex	agegrp	bp_before	bp_after
100	101	Female	60+	168	178
101	102	Female	60+	142	141
102	103	Female	60+	147	149
103	104	Female	60+	148	148
104	105	Female	60+	162	138

Reading CSV Files with Custom Delimiters

Sometimes, data in a CSV file may use a delimiter other than a comma. In that case you can use the sep or delimiter parameter to specify the custom character used as a separator.

Example

Here is an example that demonstrates specifying the custom delimiter to the sep parameter of the pandas.read_csv() method.

import pandas as pd

# Import StringIO to load a file-like object for reading CSV data
from io import StringIO

# Create a tab-delimited data
data = """
Sr.no\tName\tGender\tAge
1\tChinmayi\tfemale\t22
2\tMadhuri\tfemale\t38
3\tKarthik\tmale\t26
4\tGeetha\tfemale\t35
"""
# Use StringIO to convert the string data into a file-like object
obj = StringIO(data)

# Read a tab-separated data using read_csv()
data = pd.read_csv(obj, sep='\t')
print('Output DataFrame:')
print(data.head())

The output of the above code as follows −

Output DataFrame:

	Sr.no	Name	Gender	Age
0	1	Chinmayi	female	22
1	2	Madhuri	female	38
2	3	Karthik	male	26
3	4	Geetha	female	35

Control Reading Number of Rows from CSV

You can limit the number of rows read from a CSV file using the nrows parameter of the pandas.read_csv() method.

Example

This example demonstrates how to limit the number of rows while reading the CSV data using the pandas.read_csv() method with the nrows parameter.

import pandas as pd
from io import StringIO

# Create a CSV data
data = """
Sr.no,Name,Gender,Age
1,Chinmayi,female,22
2,Madhuri,female,38
3,Karthik,male,26
4,Geetha,female,35
"""

# Use StringIO to convert the string data into a file-like object
obj = StringIO(data)

# Read a tab-separated data using read_csv()
data = pd.read_csv(obj, nrows=2)

print('Output DataFrame:')
print(data)

Following is the output of the above code −

Output DataFrame:

	Sr.no	Name	Gender	Age
0	1	Chinmayi	female	22
1	2	Madhuri	female	38

Parsing Dates while Reading CSV Data

The parse_dates parameter handles datetime object while reading the CSV data containing the date columns.

Example

Here is an example that shows how pandas automatically parse date columns into datetime64 objects.

import pandas as pd

# Creating DataFrame 
dataFrame = pd.DataFrame({'Car': ['BMW', 'Lexus', 'Audi', 'Mercedes', 'Jaguar', 'Bentley'],
'Date_of_purchase': ['2024-10-10', '2024-10-12', '2024-10-17', '2024-10-16', '2024-10-19', '2024-10-22']})

# write dataFrame to CSV file
dataFrame.to_csv("Sample_CSV_File.csv")

# Parsing date columns
data = pd.read_csv("Sample_CSV_File.csv", parse_dates=['Date_of_purchase'])
print(data.info())

The output of the above code as follows −

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 6 entries, 0 to 5
Data columns (total 3 columns):
 #   Column            Non-Null Count  Dtype         
---  ------            --------------  -----         
 0   Unnamed: 0        6 non-null      int64         
 1   Car               6 non-null      object        
 2   Date_of_purchase  6 non-null      datetime64[ns]
dtypes: datetime64[ns](1), int64(1), object(1)
memory usage: 276.0+ bytes