Generating Numerical Sequences (arange, linspace, logspace)0%
NumPy Data Types (dtype) & Type Casting (astype)

Generating Numerical Sequences (arange, linspace, logspace)

Beginner12 min read•Updated: Jul 11, 2026
Study Materials

Generating Numerical Sequences (arange, linspace, logspace)

Mathematical modeling, data plotting, and algorithm testing often require continuous sequences of evenly distributed numbers. NumPy provides three core sequence generation engines: np.arange(), np.linspace(), and np.logspace().

1. Step-Based Sequences with np.arange()

np.arange() is the NumPy counterpart to Python's built-in range(), but returns a contiguous ndarray and supports floating-point steps:

Python
import numpy as np
 
# Syntax: np.arange([start,] stop[, step,], dtype=None)
 
# 0 to 9 (start=0 default, step=1 default)
a = np.arange(10)
print(a) # [0 1 2 3 4 5 6 7 8 9] (Stop value 10 is EXCLUDED!)
 
# 10 to 50 with step of 5
b = np.arange(10, 50, 5)
print(b) # [10 15 20 25 30 35 40 45]
 
# Floating point step
c = np.arange(0.0, 1.0, 0.2)
print(c) # [0. 0.2 0.4 0.6 0.8]
The Floating-Point Step Hazard in arange: Because computer hardware stores floating-point numbers in binary fractions, steps like 0.1 have minor precision rounding errors. Sometimes np.arange(0, 0.3, 0.1) returns 3 elements, and sometimes 4 depending on floating-point round-off! For floating-point sequences, ALWAYS use np.linspace() instead!
Array Creation and Numerical RangesClick to Zoom
Array Creation and Numerical Ranges

2. Sample-Count Sequences with np.linspace()

When generating sample points for charting (such as generating 100 points between $0$ and $2pi$ to plot a smooth Sine wave), you do not know the step size; you know how many points you want. np.linspace() solves this perfectly:

Python
# Syntax: np.linspace(start, stop, num=50, endpoint=True)
 
# Generate exactly 5 evenly spaced points between 0 and 1
pts = np.linspace(0, 1, 5)
print(pts) # [0. 0.25 0.5 0.75 1. ]

Key Properties of linspace:

  • Endpoint is Included: Unlike arange(), the stop value is included by default!
  • Calculating the Step: NumPy calculates the exact step internally: step = (stop - start) / (num - 1).
  • Retaining the Step Size: Pass retstep=True to get both the array and the calculated step delta:
Python
arr, step = np.linspace(0, 100, 5, retstep=True)
print(step) # 25.0

3. Logarithmic Scales with np.logspace()

For frequencies, decibels, seismic Richter scales, or machine learning hyperparameter tuning (e.g., testing learning rates across orders of magnitude from $10^{-4}$ to $10^{1}$):

Python
# Syntax: np.logspace(start, stop, num=50, base=10.0)
 
# Generates 5 points spaced evenly on a log scale from 10^1 (10) to 10^4 (10000)
log_pts = np.logspace(1, 4, num=4)
print(log_pts) # [ 10. 100. 1000. 10000.]
 
# Powers of 2: from 2^0 (1) to 2^8 (256)
powers_of_two = np.logspace(0, 8, num=9, base=2)
print(powers_of_two) # [ 1. 2. 4. 8. 16. 32. 64. 128. 256.]

Multiple Choice Questions

1. What will 'np.arange(2, 10, 2)' return?

A. array([2, 4, 6, 8, 10]) B. array([2, 4, 6, 8]) C. array([2, 5, 8]) D. array([4, 6, 8, 10]) Answer: B Explanation: np.arange excludes the stop boundary (10), starting at 2 and stepping by 2 to yield [2, 4, 6, 8].

2. Why is 'np.linspace()' strongly preferred over 'np.arange()' when generating floating-point ranges for data visualization?

A. linspace runs in C++ while arange runs in Python B. Floating-point rounding errors in arange can lead to unpredictable element counts and boundary inclusion C. linspace cannot handle integers D. arange does not support decimals Answer: B Explanation: Binary floating-point representation quirks can make the exact stop point in arange unpredictable; linspace guarantees the exact requested number of points.

3. By default, is the 'stop' value included or excluded in 'np.linspace(0, 10, 5)'?

A. Excluded B. Included (endpoint=True by default) C. Included only on weekends D. Rounded to nearest prime Answer: B Explanation: In np.linspace(), endpoint=True by default, meaning the stop value (10) is guaranteed to be the final element of the array.

4. What will 'np.linspace(0, 100, 5)' produce?

A. array([0., 20., 40., 60., 80.]) B. array([0., 25., 50., 75., 100.]) C. array([0., 10., 20., 30., 40.]) D. array([5., 25., 50., 75., 100.]) Answer: B Explanation: Spacing 5 numbers evenly between 0 and 100 divides the interval into 4 steps of 25.0: 0, 25, 50, 75, 100.

5. What does the first argument 'start=2' represent in 'np.logspace(2, 5, num=4, base=10)'?

A. The number 2 B. The exponent power 10^2 = 100 C. 20 D. A 2-dimensional matrix Answer: B Explanation: In logspace, the start and stop arguments represent exponents of the specified base (10^2 to 10^5).

Next Lesson

NumPy Data Types (dtype) & Type Casting (astype)

Continue learning with hands-on practice, examples, and exercises in the upcoming topic.

Related Lessons

Previous LessonNext Lesson
Built-in Creation Routines (zeros, ones, full, empty)NumPy Data Types (dtype) & Type Casting (astype)

Practice Quiz

Test your understanding of this lesson with 5 questions. Each question has one correct answer.

PreviousNext
PrevNext