How to read CSV data into a record array in NumPy?

In NumPy, a record array (or recarray) is a specialized array that allows you to access fields as attributes, providing a convenient way to handle structured data. It is essentially a structured array that is wrapped in an object-oriented interface, making it easier to work with data that has named columns, similar to a DataFrame in pandas.

Key Features of Record Arrays

1. Named Fields: Each column in a record array can have a name, allowing you to access it using attribute-style access (e.g., array.fieldname).
2. Mixed Data Types: Different fields can have different data types (e.g., integers, floats, strings).
3. Convenient Access: You can access and modify the fields by their names, similar to how you work with attributes in objects.

Read CSV data into a record array in NumPy

• Using numpy.genfromtxt
• Using recfromcsv function
• Using pandas and converting to NumPy record array

Read CSV data into a record array in NumPy Using numpy.genfromtxt

In this approach example, we are using numpy.genfromtxt to read CSV data (geeks_data.csv) into a NumPy record array (data). The delimiter=',' specifies that fields are separated by commas, dtype=None allows NumPy to infer data types, names=True indicates that the first row contains column names, and encoding=None handles the default encoding.

# Approach 1: Using numpy.genfromtxt 

import numpy as np

csv_data = """
ID,Name,Salary
1,Gaurav,50000
2,Yuvraj,60000
3,Pranav,55000
"""
with open('geeks_data.csv', 'w') as f:
    f.write(csv_data)
    
data = np.genfromtxt('geeks_data.csv', delimiter=',', 
                     		dtype=None, names=True, encoding=None)

print("Record array:")
print(data)

Record array:
[(1, 'Gaurav', 50000) (2, 'Yuvraj', 60000) (3, 'Pranav', 55000)]

Read CSV data into a record array in NumPy Using recfromcsv function

In this approach using np.recfromcsv, we directly read the CSV data (geeks_data.csv) into a NumPy record array (data). The function np.recfromcsv is specifically designed to handle CSV files and creates a structured array where each column can be accessed using its name.

# Approach 2: Using recfromcsv function 
import numpy as np
csv_data = """
ID,Name,Salary
1,Gaurav,50000
2,Yuvraj,60000
3,Pranav,55000
"""
with open('geeks_data.csv', 'w') as f:
    f.write(csv_data)
data = np.recfromcsv('geeks_data.csv', encoding=None)
print("Record array:")
print(data)

Record array:
[(1, 'Gaurav', 50000) (2, 'Yuvraj', 60000) (3, 'Pranav', 55000)]

Read CSV data into a record array in NumPy Using pandas and converting to NumPy record array

In this approach example, we are using pandas to read the CSV data (geeks_data.csv) into a DataFrame (df). The read_csv function parses the CSV file, and then to_records() converts the DataFrame into a NumPy record array (data). Setting index=False ensures that the DataFrame index is not included in the resulting record array.

# Approach 3: Using pandas and converting to NumPy record array

import numpy as np
import pandas as pd
csv_data = """
ID,Name,Salary
1,Gaurav,50000
2,Yuvraj,60000
3,Pranav,55000
"""
with open('geeks_data.csv', 'w') as f:
    f.write(csv_data)
df = pd.read_csv('geeks_data.csv')
data = df.to_records(index=False)
print("Record array:")
print(data)

Record array:
[(1, 'Gaurav', 50000) (2, 'Yuvraj', 60000) (3, 'Pranav', 55000)]

Conclusion

In conclusion, parsing CSV data into NumPy record arrays using numpy.genfromtxt, np.recfromcsv, or pandas provides flexible options for handling structured data in Python. These approaches allow seamless integration of CSV data into NumPy arrays, enabling efficient data manipulation and analysis tasks within data science and numerical computing workflows.