Shannon Entropy Calculator

Calculate information entropy to measure randomness and uncertainty in probability distributions

Calculate Shannon Entropy

Probabilities

Current sum: 1.0000

Shannon Entropy Results

1.0000
Entropy (bits (shannons))
1.0000
Maximum Entropy
100.0%
Relative Entropy

Very high entropy - maximum randomness achieved

Shannon Entropy Formula: H(X) = -∑ P(xi) × logb(P(xi))

Base: 2

Events: 2

Example Calculation

Sequence Analysis Example

Sequence: 1 0 3 5 8 3 0 7 0 1

Character frequencies:

  • • '1': 2 occurrences (probability = 2/10 = 0.2)
  • • '0': 3 occurrences (probability = 3/10 = 0.3)
  • • '3': 2 occurrences (probability = 2/10 = 0.2)
  • • '5': 1 occurrence (probability = 1/10 = 0.1)
  • • '8': 1 occurrence (probability = 1/10 = 0.1)
  • • '7': 1 occurrence (probability = 1/10 = 0.1)

Calculation (Base 2)

H = -∑ P(xi) × log₂(P(xi))

H = -[0.2×log₂(0.2) + 0.3×log₂(0.3) + 0.2×log₂(0.2) + 0.1×log₂(0.1) + 0.1×log₂(0.1) + 0.1×log₂(0.1)]

H = -[0.2×(-2.32) + 0.3×(-1.74) + 0.2×(-2.32) + 0.1×(-3.32) + 0.1×(-3.32) + 0.1×(-3.32)]

H ≈ 2.446 bits

Entropy Units

2

Bits (Shannons)

Base 2 logarithm

Most common in computing

e

Nats

Natural logarithm (base e)

Common in mathematics

10

Dits/Bans/Hartleys

Base 10 logarithm

Historical significance

Entropy Tips

Maximum entropy occurs when all events are equally likely

Zero entropy means perfect predictability (one outcome)

Higher entropy indicates more randomness and unpredictability

Probabilities must sum to 1.0 for valid entropy calculation

Use frequencies when you have count data instead of probabilities

Understanding Shannon Entropy

What is Shannon Entropy?

Shannon entropy, developed by Claude Shannon in 1948, is a measure of the average amount of information or uncertainty contained in a message or system. It quantifies how unpredictable or random the outcomes of a probability distribution are.

Applications

  • Information Theory: Data compression and communication
  • Ecology: Biodiversity and species distribution analysis
  • Machine Learning: Decision trees and feature selection
  • Cryptography: Password strength and randomness testing

Formula Explanation

H(X) = -∑ P(xi) × logb(P(xi))

Shannon Entropy Formula

  • H(X): Shannon entropy of random variable X
  • P(xi): Probability of outcome xi
  • logb: Logarithm with base b
  • ∑: Sum over all possible outcomes

Note: The negative sign ensures entropy is always non-negative, since log(P(xi)) is negative when 0 ≤ P(xi) ≤ 1.

Information Content

Measures the average information content per symbol in a message or data stream.

Uncertainty

Quantifies the unpredictability of outcomes in a probability distribution.

Diversity Index

Used in ecology and other fields to measure diversity and evenness of distributions.