Python math.atanh(): Inverse Hyperbolic Tangent

Syntax & Parameters 🔝

The math.atanh() function is part of the standard math module:

import math

result = math.atanh(x)

Return Value: Returns a floating-point number (float) spanning (-∞, +∞).

Basic Examples 🔝

Evaluating math.atanh() for positive, zero, and negative values:

import math

print("atanh(0):   ", math.atanh(0))      # 0.0
print("atanh(0.5): ", math.atanh(0.5))    # 0.5493061443340548
print("atanh(-0.5):", math.atanh(-0.5))   # -0.5493061443340548 (odd symmetry)
print("atanh(0.95):", math.atanh(0.95))   # 1.8317808204646738

atanh(0):    0.0
atanh(0.5):  0.5493061443340548
atanh(-0.5): -0.5493061443340548
atanh(0.95): 1.8317808204646738

Logarithmic Definition & Equivalence 🔝

The inverse hyperbolic tangent is mathematically defined through the natural logarithm:

You can verify this identity directly in Python:

import math

x = 0.6
val_builtin = math.atanh(x)
val_formula = 0.5 * math.log((1 + x) / (1 - x))

print("math.atanh(0.6):", val_builtin)
print("Log formula:     ", val_formula)
print("Is close?        ", math.isclose(val_builtin, val_formula))  # True

Open Domain Restrictions: (-1.0, 1.0) 🔝

Unlike circular inverse trigonometry where endpoints are valid (e.g. arcsin(1.0) = π/2), the domain of math.atanh() is strictly an open interval:

• At x = 1.0, the denominator 1 - x becomes zero and the logarithm approaches +∞.
• At x = -1.0, the numerator 1 + x becomes zero and the logarithm approaches -∞.
• For |x| ≥ 1.0, Python raises a ValueError: math domain error.

import math

invalid_inputs = [-1.5, -1.0, 1.0, 2.0]

for val in invalid_inputs:
    try:
        math.atanh(val)
    except ValueError as err:
        print(f"atanh({val}) raises: {err}")

atanh(-1.5) raises: math domain error
atanh(-1.0) raises: math domain error
atanh(1.0) raises: math domain error
atanh(2.0) raises: math domain error

math.atan() vs. math.atanh(): Circular vs. Hyperbolic 🔝

Learners often confuse circular arc tangent (math.atan) with hyperbolic arc tangent (math.atanh). Here is how their domains, outputs, and geometric meanings differ:

Plotting the Inverse Hyperbolic Curve with Matplotlib 🔝

Visualizing the asymptotic rise toward vertical asymptotes at x = ±1 using Matplotlib:

import math
import matplotlib.pyplot as plt

x_vals = [i * 0.02 - 0.98 for i in range(0, 99)]  # -0.98 to +0.98
y_vals = [math.atanh(x) for x in x_vals]

plt.plot(x_vals, y_vals, label="atanh(x)", color="#1a73e8")
plt.axvline(1.0, color="red", linestyle="--", label="x = 1 asymptote")
plt.axvline(-1.0, color="red", linestyle="--", label="x = -1 asymptote")
plt.title("Inverse Hyperbolic Tangent Curve")
plt.xlabel("Input x in (-1, 1)")
plt.ylabel("atanh(x)")
plt.grid(True)
plt.legend()
plt.show()

Hands-on Exercise: Fisher z-Transformation for Correlation with Assertions 🔝

In statistics, Pearson correlation coefficients r are bounded between -1.0 and +1.0, which causes sample correlations near ±1 to follow a skewed distribution. Fisher's z-transformation transforms r into an approximately normal distribution using atanh():

import math

def fisher_z_transform(correlation_r):
    if abs(correlation_r) >= 1.0:
        raise ValueError("Pearson r must be strictly between -1.0 and 1.0.")
        
    z_score = math.atanh(correlation_r)
    return round(z_score, 4)

# Unit Assertions
assert fisher_z_transform(0.0) == 0.0       # Zero correlation maps to z=0
assert fisher_z_transform(0.5) == 0.5493    # Moderate positive correlation
assert fisher_z_transform(-0.5) == -0.5493  # Moderate negative correlation
assert fisher_z_transform(0.75) == 0.9730   # Strong correlation
assert fisher_z_transform(0.90) == 1.4722   # Very strong correlation

print("Fisher z for r=0.75:", fisher_z_transform(0.75))
print("Fisher z for r=0.90:", fisher_z_transform(0.90))
print("All Fisher z-transformation assertions passed successfully!")

Practice in Google Colab & GitHub 🔝

Open in Google Colab View on GitHub

Run all math.atanh() examples interactively, test domain bounds, verify logarithmic formula equivalence, and calculate Fisher z-transformations directly in Google Colab.

More Details & Related Guides 🔝

• Python Math Module Overview — Master arithmetic, trigonometry, and numeric constants in Python.
• math.tanh() Function — Forward hyperbolic tangent function.
• math.asinh() Function — Inverse hyperbolic sine of unrestricted real numbers.
• math.acosh() Function — Inverse hyperbolic cosine for values ≥ 1.0.
• math.atan() Function — Circular inverse arc tangent returning radians in (-π/2, π/2).
• math.log() Function — Natural logarithm and foundation of hyperbolic inverses.
• Exception Handling in Python — Managing ValueError and math domain errors gracefully.

Subhendu Mohapatra

Author & Instructor at plus2net

I write and maintain practical tutorials on Python, PHP, SQL, JavaScript, HTML, jQuery, and web development at plus2net. The tutorials focus on clear explanations, working examples, and code that readers can test and adapt while learning.

About the author   Watch on YouTube

Connect: