---
title: Bandt–Pompe Symbolization Methodology
url: https://www.emergentmind.com/topics/bandt-pompe-symbolization-methodology-4b91630f-c536-434b-b404-c6656aace16e
type: topic
---

# Bandt–Pompe Symbolization Methodology

The Bandt–Pompe symbolization methodology is a rank-based approach for transforming real-valued time series into symbolic sequences of ordinal patterns, facilitating the quantitative characterization of complex dynamics using information-theoretic measures such as permutation entropy and statistical complexity. Ordinal analysis, as introduced by Bandt and Pompe in 2002, has become a foundational tool in the study of time series from diverse domains, leveraging the temporal ordering relations within embedded vectors to yield robust, easily interpretable statistics that are insensitive to monotonic transformations and measurement noise.

## 1. Formal Definition and Symbolization Procedure

The methodology begins with the construction of overlapping delay vectors from a given real-valued time series $X = \{x_t\}_{t=1}^N$, using two key parameters: embedding dimension $D \geq 2$ and temporal delay $\tau \geq 1$ [1510.04210][1601.06925][2204.12883][2411.13761][2508.19955]. For $t$ ranging from $1$ to $N-(D-1)\tau$,

\[
\mathbf{X}_t = [x_t,\, x_{t+\tau},\, \ldots,\, x_{t+(D-1)\tau}]
\]

Each vector is mapped to an ordinal pattern (permutation) $\pi = (r_0, r_1, ..., r_{D-1})$ such that

\[
x_{t+r_0\tau} \leq x_{t+r_1\tau} \leq \cdots \leq x_{t+r_{D-1}\tau}
\]

Ties are resolved by prescribing that earlier time indices are ranked lower (i.e., $r_i < r_{i+1}$ if $x_{t + r_i \tau} = x_{t + r_{i+1} \tau}$) [1601.06925][2204.12883][2411.13761].

The mapping of all such vectors yields a sequence of symbols (ordinal patterns) taking as values the $D!$ possible permutations of $D$ indices.

## 2. Permutation Probability Distribution and Information-Theoretic Quantifiers

The empirical distribution of observed patterns $P(\pi)$ is defined as

\[
p(\pi) = \frac{\#\{\text{vectors assigned to } \pi\}}{N-(D-1)\tau}
\]

Permutation entropy $S[P]$ quantifies the diversity of observed patterns via the Shannon entropy [1510.04210][1601.06925][2411.13761][2508.19955]:

\[
S[P] = -\sum_{\pi} p(\pi) \log p(\pi)
\]
\[
H[P] = \frac{S[P]}{\log D!} \in [0,1]
\]

To capture both randomness and structure, statistical complexity $C[P]$ is constructed by combining normalized entropy with a disequilibrium measure such as the normalized Jensen–Shannon divergence $Q_J[P, P_e]$ relative to the uniform distribution $P_e(\pi) = 1/D!$:

\[
J[P, P_e] = S\left(\frac{P+P_e}{2}\right) - \frac{1}{2}S[P] - \frac{1}{2}S[P_e]
\]
\[
Q_J[P,P_e] = \frac{J[P,P_e]}{J_{\max}}
\]
\[
C[P] = H[P] \times Q_J[P,P_e]
\]

This construction ensures $C[P] \in [0,1]$ and is maximized for distributions that are neither completely random nor perfectly ordered [1510.04210][1601.06925][2508.19955]. Alternative complexity quantifiers, such as LMC complexity and, for discrete distributions, Fisher information, are also widely employed [1601.06925][1304.0399].

## 3. Practical Considerations: Parameter Selection and Tie Handling

Selection of the embedding dimension $D$ and delay $\tau$ is informed by trade-offs between pattern diversity and statistical reliability. Bandt and Pompe recommend $3 \leq D \leq 7$, with $N \gg D!$ (e.g., $N \gtrsim 5D!$) to ensure all patterns are well-sampled [1510.04210][1601.06925][2204.12883][2411.13761][2508.19955]. The time delay $\tau$ is often set to $1$ for consecutive samples but can be increased to probe slower dynamics or specific time scales indicated by autocorrelation or mutual information analyses [1510.04210][2411.13761].

Ties in floating-point data are rare, but real-world measurement constraints (e.g., quantization or plateaus) can produce equal values. The canonical approach resolves ties by temporal order—earlier indices are ranked lower [2204.12883][2411.13761][1510.04210]. In high-frequency tied data such as HRV time series, alternative tie-breaking or imputation strategies are required. Bayesian missing data imputation, which infers a tie-case's permutation based on observed pattern frequencies from complete cases, yields minimally biased entropy estimates [1707.01517].

## 4. The Entropy–Complexity Plane and Dynamical Inference

Plotting normalized permutation entropy $H[P]$ versus statistical complexity $C[P]$ situates time series within a causality plane whose boundaries are determined by theoretical minima and maxima of $C$ at each entropy level [1510.04210][2204.12883][2411.13761][2511.09243]. Different regimes cluster in distinct regions:

- Purely stochastic (white noise): $H[P] \to 1$, $C[P] \to 0$
- Deterministic chaos: intermediate $H[P]$, maximal $C[P]$
- Regular deterministic: $H[P] \to 0$, $C[P] \to 0$

This framework robustly distinguishes stochastic, chaotic, and regular dynamics, and is resilient to noise contamination. Analyses of forbidden (never occurring in deterministic systems) and missing (unobserved due to finite sampling) patterns further decode the presence of underlying deterministic structure, even under strong noise [1110.0776][1802.01496].

## 5. Extensions: Weighted, Global, and Multivariate Variants

Weighted permutation entropy (WPE) enhances classical permutation entropy by incorporating amplitude information, such as local variance within each embedded vector, resulting in a weighted distribution over ordinal patterns [2511.09243]. WPE improves sensitivity to changes in both ordering and amplitude dynamics.

Global Permutation Entropy (GPE) generalizes the standard consecutive-pattern PE by considering all possible order-$d$ patterns, including non-consecutive point subsets, leveraging efficient algorithms to scale the enumeration [2508.19955]. GPE reveals structural information otherwise inaccessible through the classical consecutive embedding approach.

Multivariate generalizations, bivariate and network extensions facilitate ordinal pattern analysis of multidimensional or coupled time series. Permutation mutual information and transfer entropy derived from joint or conditional pattern distributions are increasingly used to probe coupling and causality in networked systems [2301.09566].

## 6. Representative Applications and Impact

Bandt–Pompe analysis has demonstrated value across physics, physiology, neuroscience, geosciences, and engineering. Applications include:

- Biomedical signal analysis: EEG and iEEG-based separation of cognitive or sleep states [2511.09243][2507.16159][2101.08905], HRV analyses for cardiological assessment [1707.01517], detection of seizures [2204.12883], or classification in clinical neuroscience [2301.09566].
- Dynamical system analysis: Distinguishing stochasticity from determinism in cardiac arrhythmia [1802.01496], detecting forbidden patterns and determinism under noise [1110.0776], inferring macroscopic behaviors in complex systems (e.g., vehicle traffic, neural activity) [1510.04210][1304.0399].
- Nonlinear signal processing: Feature extraction for classification, clustering, and model identification, including hybrid approaches that combine BP symbolic features with machine learning pipelines [1601.06925][2508.19955][2204.12883][2507.09381].

A condensed table of core metric definitions is shown below (notation as in text):

| Quantifier               | Formula                                                                                   | Range        |
|--------------------------|------------------------------------------------------------------------------------------|--------------|
| Permutation entropy      | $H[P] = -\sum_\pi p(\pi) \log p(\pi)/\log D!$                                            | $[0,1]$      |
| Jensen–Shannon divergence| $J[P, P_e] = S[(P+P_e)/2] - \frac{1}{2}S[P] - \frac{1}{2}S[P_e]$                         | $[0, J_{max}]$|
| Statistical complexity   | $C[P] = H[P] \cdot Q_J[P,P_e]$                                                           | $[0,1]$      |

## 7. Open Problems and Limitations

Current limitations include the statistical characterization of ordinal measures under non-stationarity and in finite samples, principled feature reduction for ordinal-pattern-based machine learning, reliable handling of missing or irregularly sampled data, and extensions to higher-dimensional or network time series [2204.12883][2301.09566]. While the methodology is robust to monotonic transformations and moderate noise, it discards amplitude magnitudes, which may limit its discriminative power in settings where both order and value are informative [2507.09381][2204.12883].

Combining Bandt–Pompe features with kernel or deep-learning models is an active area of research, promising further improvements in time series classification and forecasting [2204.12883][1601.06925]. Ongoing studies explore optimal parameterization, efficient computation for high $D$, and integration with non-parametric or compression-based complexity measures [1310.1379][1707.05164].

---

**References**:  
[1510.04210], [1601.06925], [2204.12883], [2411.13761], [2508.19955], [1110.0776], [2507.09381], [1802.01496], [1304.0399], [1707.05164], [2507.16159], [1310.1379], [2301.09566], [1502.01309], [2101.08905], [1707.01517], [2511.09243].

Source: https://www.emergentmind.com/topics/bandt-pompe-symbolization-methodology-4b91630f-c536-434b-b404-c6656aace16e