MSK Institute • Shikohabad
Loading Page Content...
Preparing courses, live labs, and learning resources.सामग्री लोड हो रही है, कृपया प्रतीक्षा करें...
Connecting serverPlease wait...
💻 Practical Labs🚀 Free Demo📜 Verified Certificates
Universal Functions (ufuncs), Trigonometry & Logarithms0%
Universal Functions (ufuncs), Trigonometry & Logarithms
Beginner12 min read•Updated: Jul 11, 2026
Mathematical ufuncs: Trigonometric, Exponential, Logarithmic
At the technical core of NumPy's blazing numerical speed lies the concept of Universal Functions (commonly abbreviated as ufuncs).
A ufunc is a function that operates on ndarrays in an element-by-element fashion. Unlike standard Python math functions from the math module (which accept only single scalar numbers and must be invoked sequentially inside loops), NumPy ufuncs are compiled C routines that support broadcasting, type coercion, and direct memory output targeting.
Click to Zoom1. Trigonometric and Angle Routines
NumPy provides complete implementations of trigonometric functions. All trigonometric ufuncs expect angles expressed in radians (not degrees):
$$\text{radians} = \text{degrees} \times \frac{\pi}{180}$$
import numpy as np
# Create angles in degrees and convert to radians
angles_deg = np.array([0, 30, 45, 60, 90, 180, 270, 360])
angles_rad = np.deg2rad(angles_deg) # or np.radians(angles_deg)
# Evaluate sine, cosine, and tangent simultaneously
sin_values = np.sin(angles_rad)
cos_values = np.cos(angles_rad)
tan_values = np.tan(angles_rad)
print("Degrees: ", angles_deg)
print("Sine: ", np.round(sin_values, 4))
print("Cosine: ", np.round(cos_values, 4))
Output:
Degrees: [ 0 30 45 60 90 180 270 360]
Sine: [ 0. 0.5 0.7071 0.866 1. 0. -1. -0. ]
Cosine: [ 1. 0.866 0.7071 0.5 0. -1. -0. 1. ]
Inverse Trigonometric Functions and arctan2:
When computing angles from Cartesian coordinates $(x, y)$, standard arctan(y/x) loses quadrant information and risks division by zero. Always use np.arctan2(y, x):
y = np.array([1.0, 1.0, -1.0, -1.0])
x = np.array([1.0, -1.0, -1.0, 1.0])
# Correct quadrant angle in [-pi, pi]
theta = np.rad2deg(np.arctan2(y, x))
print("Quadrant angles (deg):", theta) # [45. 135. -135. -45.]
2. Exponential and Logarithmic ufuncs
Logarithmic and exponential transformations are essential across data preprocessing (log-transforming skewed financial returns or heavy-tailed distributions):
x = np.array([1.0, 10.0, 100.0, 1000.0])
# Natural logarithm (base e)
print("Natural log (ln x): ", np.log(x))
# Common logarithm (base 10)
print("Base-10 log (log10 x): ", np.log10(x))
# Binary logarithm (base 2 - information theory / bits)
print("Base-2 log (log2 x): ", np.log2(x))
# Natural exponential (e^x)
powers = np.array([0.0, 1.0, 2.0, 3.0])
print("exp(x): ", np.exp(powers))
Numerical Precision: log1p and expm1
When $x$ is near zero, evaluating $\ln(1 + x)$ using np.log(1 + x) suffers from catastrophic floating-point cancellation. NumPy provides np.log1p(x) and its inverse np.expm1(x):
tiny = 1e-15
print("Standard log(1 + tiny):", np.log(1.0 + tiny)) # loses precision: 0.0 or truncated
print("Precision log1p(tiny): ", np.log1p(tiny)) # accurate: 1e-15
3. High-Performance Optimization: The out Parameter
Every time you execute y = np.sin(x), NumPy allocates a brand new memory buffer for the result. When processing massive datasets (gigabytes of sensor streams), continuous memory allocations trigger garbage collection and waste bandwidth.
All ufuncs provide an optional out parameter to write results directly into an existing pre-allocated buffer:
data = np.linspace(0, 10, 10_000_000)
output_buffer = np.empty_like(data)
# Writes directly into output_buffer without allocating new RAM!
np.sin(data, out=output_buffer)
print("Buffer successfully populated in-place. Mean:", output_buffer.mean())
Interactive Quiz
Multiple Choice Questions
Question 1
In what unit of measurement do trigonometric ufuncs like
np.sin() and np.cos() expect input angles?Question 2
Why is
np.arctan2(y, x) preferred over np.arctan(y / x) in computer vision and robotics?Question 3
What is the benefit of using
np.log1p(x) instead of np.log(1 + x) when x is extremely small (e.g. 1e-15)?Question 4
What does the
out parameter in a universal function like np.exp(x, out=buffer) accomplish?Question 5
Which NumPy ufunc computes the common base-10 logarithm?
Next Lesson
Aggregate Statistics (sum, mean, median, std, var)
Continue learning with hands-on practice, examples, and exercises in the upcoming topic.
Related Lessons
| Previous Lesson | Next Lesson |
|---|---|
| The Rules of NumPy Broadcasting Explained | Aggregate Statistics (sum, mean, median, std, var) |
Practice Quiz
Test your understanding of this lesson with 5 questions. Each question has one correct answer.