EconBase
← Back to paper

Multi-Band Variable-Lag Granger Causality: A Unified Framework for Causal Time Series Inference across Frequencies

Extracted main text — title through conclusion, appendix excluded. This is what our citation measures are computed over, published so the extraction can be checked by eye.

41,493 characters · 25 sections · 27 citation commands

Rendered from LaTeX for readability, not typeset faithfully. Citation keys are highlighted; maths is left as source; figures, tables and equation environments are summarised rather than reproduced; unrecognised commands are greyed out so nothing is silently dropped. Email addresses are removed.

My Publication Title --- Multiple Authors

abstractUnderstanding causal relationships in time series is fundamental to many domains, including neuroscience, economics, and behavioral science. Granger causality is one of the well-known techniques for inferring causality in time series. Typically, Granger causality frameworks have a strong fix-lag assumption between cause and effect, which is often unrealistic in complex systems. While recent work on variable-lag Granger causality (VLGC) addresses this limitation by allowing a cause to influence an effect with different time lags at each time point, it fails to account for the fact that causal interactions may vary not only in time delay but also across frequency bands. For example, in brain signals, alpha-band activity may influence another region with a shorter delay than slower delta-band oscillations. In this work, we formalize Multi-Band Variable-Lag Granger Causality (MB-VLGC) and propose a novel framework that generalizes traditional VLGC by explicitly modeling frequency-dependent causal delays. We provide a formal definition of MB-VLGC, demonstrate its theoretical soundness, and propose an efficient inference pipeline. Extensive experiments across multiple domains demonstrate that our framework significantly outperforms existing methods on both synthetic and real-world datasets, confirming its broad applicability to any type of time series data. Code and datasets are publicly available.

Introduction

figure*[figure* omitted — 370 chars of source]

Understanding causal relationships in time series data is fundamental across numerous scientific disciplines, including neuroscience, economics, and behavioral science. One of important questions in these domains is identifying which time-dependent signals initiate or influence other patterns of behavior over time. Experimental methods such as randomized controlled trials, in many cases, are infeasible due to ethical, logistical, or financial constraints varian2016causal. Consequently, methods for inferring causality from observational time series data have become crucial roles for scientific inquiry.

One of the well-known frameworks for such analysis is Granger causality granger1969investigating. It defines a directional causal influence from a time series $X$ to another time series $Y$ if the inclusion of $X$'s past improves the prediction of $Y$ beyond the predictive power of $Y$'s own past. Despite its popularity, traditional Granger causality methods rely on a critical assumption: the time delay between cause and effect is fixed. However, in many phenomena, the delay between a cause and its effect may vary over time, rendering the fixed-lag assumption overly restrictive.

To address this issue, the concept of Variable-Lag Granger Causality (VLGC) 10.1145/3441452 was introduced. VLGC allows for temporal flexibility by accommodating time-varying delays in causal relationships using Dynamic Time Warping (DTW) 1163055. Nevertheless, VLGC overlooks a crucial aspect of frequency structures in signals. For example, in neuroscience, causal effects often vary across frequency bands canolty2010functional; for instance, alpha-band oscillations in the brain may exert influence more rapidly than slower delta-band activity michalareas2016alpha.

Hence, in this work, we formalize Multi-Band Variable-Lag Granger Causality (MB-VLGC), a unified framework that generalizes VLGC by modeling frequency-specific time delays. In addition to standard Granger causality and variable-lag methods, our framework offers:

itemize• Unified time-frequency causal inference: MB-VLGC extends both Variable-Lag Granger Causality (VLGC) and frequency-domain GC by combining spectral decomposition with dynamic temporal alignment. • Frequency-specific delay modeling: Our framework infers causal interactions with distinct time lags across multiple frequency bands.

MB-VLGC reveals multiscale causal structures and offers a general tool for analyzing complex time series across scientific domains.

Related works

Granger causality (GC) is a widely adopted statistical framework for time series causality, establishing that a signal $X$ is causal to $Y$ if the past of $X$ improves the prediction of $Y$ beyond what $Y$’s own past can provide granger1969investigating. While GC is widely used in economics and science, GC is simple, which make it have many limitations.

There were many methods were developed to overcome several issues of GC. For instance, since GC does not indicate how an influence may be distributed across frequencies, frequency-domain extensions of GC, such as Geweke’s spectral causality Geweke1982 and Hosoya’s decomposition hosoya1991decomposition, were introduced and they enable analysis of frequency-specific causal interactions. For linearity aspect of GC, methods (e.g. Transfer Entropy (TE) Schreiber2000,BEHRENDT2019100265, PCMCI runge2020discovering) were developed to capture non-linear dependencies. Nevertheless, all methods mentioned above relied on fixed-lag conditionals; they lack the capacity to explicitly model time-varying delays.

Regarding of time-varying delays, in many complex systems, the delay of influence can vary over time, violating this fixed-lag assumption. To relax this constraint, the work in 10.1145/3441452 proposed the concept of Variable-Lag Granger Causality (VLGC) and Variable-Lag Transfer Entropy (VLTE), which allow flexible, time-varying delays by aligning time series via DTW. VLGC and VLTE generalize classical Granger causality by permitting different lag lengths at different times, thereby capturing causal relationships that shift in time. Both approaches are also able to deal with non-stationary and unstable-causal-structure time series\footnote{Given $X$ causes $Y$, the unstable-causal-structure occurs when $\exists t\neq t'$ and $Y_t=f(X), Y_{t'}=f'(X)$ s.t. $ f\neq f'$}. However, existing GC methods (including VLGC and frequency-domain approaches) do not account for frequency-specific variations in causal influence.

In other words, all previous works either infer variable lags or frequency-specific effects, but not both. Hence, we propose MB-VLGC to fill this gap by unifying dynamic temporal alignment with spectral decomposition, enabling accurate causal inference across time and frequency. \newtheorem{definition}{Definition} \newtheorem{theorem}{Theorem}[section] \newtheorem{lemma}[theorem]{Lemma} \newtheorem{proposition}[theorem]{Proposition} \newtheorem{corollary}[theorem]{Corollary}

\newenvironment{proof}[1][Proof]{

trivlist• }{

} \newenvironment{example}[1][Example]{

trivlist• }{

} \newenvironment{remark}[1][Remark]{

trivlist• }{

}

Multi-Band Variable-Lag Granger Causality formalization

definition{Granger causality under stationary and VAR assumption}\\ Let $U_t:=[X_t,Y_t]^T$ be a zero-mean, stationary, multivariate VAR process: $U_t = \sum_{k=1}^{p}{A_kU_{t-k}}+\varepsilon_t$, where $\varepsilon_t\sim iid(0,\Sigma) $ Let $G(z)$ be a stable, invertible multivariate linear filter with no cross-component coupling. Then for all $\omega \in [0,\pi]$, the spectral Granger causality from $Y\rightarrow X$ is invariant under filtering: $f_{Y\rightarrow X}(\omega)=f_{\tilde{Y}\rightarrow \tilde{X}}(\omega)$

As shown in Barnett (2011), under the assumptions of a stationary and a VAR, Granger causality is invariant under filtering. Given that filtering does not facilitate the separation of Granger causality by frequency, as described in Definition 1. In this work, we relax the assumptions of stationary and VAR, and investigate the properties of Granger causality under filtering.

propositionIf the stationary and/or VAR assumption is dropped, then in general $f_{Y\rightarrow X}(\omega)\neq f_{\tilde{Y}\rightarrow \tilde{X}}(\omega)$: cannot conclude that the filtering does not help to separate Granger causality by frequency.
proofLet $X_t := X_{t-1} + \varepsilon_t^x$, $Y_t := \varepsilon^y_t$ where $\varepsilon_t^x,\varepsilon^y_t \sim iid\mathcal{N}(0,1)$ This is a nonstationary process, not representable by a stationary VAR process with properties $E[X_t] = E[X_{t-1}] + E[\varepsilon_t^x]=E[X_{t-1}]$ which is constant and $Var(X_t) = Var(X_{t-1}) +1$ then $Var(X_t)=t$. \\Define the filter $G(L):=1-L$, where $L$ is a lag operator, which is linear, time invariant, diagonal, stable, and invertible in $\mathbb{Z}/\{1\}$ and define $\tilde{X_t}:= G(L)X_t=X_t-X_{t-1}=\varepsilon^x_t$ and $\tilde{Y_t} :=G(L)Y_t=\varepsilon_t^y-\varepsilon^y_{t-1}$ So: $\tilde{x_t}\sim iid\mathcal{N}(0,1)$ and $\tilde{y_t}\sim MA(1)$. We knew from the granger1974spurious that $f_{Y \rightarrow X}(\omega)>0$ from a spurious Granger causality. And from $Cov(\tilde{X_t},\tilde{Y}_{t-k})=0, \forall k \geq 1$ makes $f_{\tilde{Y}\rightarrow \tilde{X}}(\omega)=0$. So, we cannot conclude that filtering does not help separate Granger-causality by frequency.

According to Proposition 2.1, when the assumptions of stationary and a VAR representation are relaxed, filtering may facilitate the separation of Granger causality by frequency. To investigate this, we use the Variable-Lag Granger Causality (VLGC) method described in Definition 2, using a base case in which the signals are decomposed into separate frequency bands: $X = X_1 + X_2$ and $Y = Y_1 + Y_2$. To validate the reliability of the VL-Granger causality test on these separated bands, we analyze the associated residual variances, as discussed in Proposition 2.2.

definition{VL-Granger Causality} \\ For time series $X$ and $Y$, and an upper bound of time lag on any two time series to be tested or the maximum time lag $ \delta_{max} \in \mathbb{N}$. The residual of the regression $r^*_{YX}$ can be defined as follow: $$r^*_{YX}(t)=Y(t)-\sum^{\delta_{max}}_{i=1}(a_iY(t-i)+b_iX(t-i)+c_iX^*(t-i))$$ Where $X^*(t-i) = X(t-i+1-\Delta_{t-i+1})$ with a time delay parameter determined by the optimal alignment path $P^*$ that minimize the regression residual between $X$ and$Y$, $\Delta_t > 0$. The coefficients $a_i,b_i$ and $c_i$ are estimated such that the residuals $r_Y,r_{YX},$ and $r^*_{YX}$ are minimized. We can conclude that $X$ VL-Granger causes $Y$ if and only if $Var(r^*_{YX})$ is less than both $Var(r_{Y})$ and $Var(r_{YX})$.

Next, we extend the concept of VLGC in Def. (ref) to work on multiple frequency bands.

definition{Multi-Band Variable-Lag Granger Causality (MB-VLGC)} Given $X,Y$ and a set of interval bands $\mathcal{B}=\{B_1,\dots,B_k\}$ s.t. $Bi=[\omega,\omega']$ is a frequency band start at $\omega$ and end at $\omega'$. We say $X$ Multi-Band Variable-Lag Granger causes $Y$ if $\exists B_i\in\mathcal{B}$, $X^{(B_i)}$ VL-Granger causes $Y^{(B_i)}$ where $(X^{(B_i)},Y^{(B_i)})$ is a pair of signals band-limited at band $B_i$.

In the next preposition, we show that the variance of residuals of MB-VLGC has the variance of residuals of traditional VLGC as its upper bound and it can be lower than VLGC's for some cases.

propositionLet $r^*_{(i)}$ be the residual from fitting $Y^{(i)} \sim VL-G(X^{(i)};P^{(i)})$ and $r^*$ be the residual from fitting a single VL-Granger model to $(X,Y)$, with alignment $P$. Then, $Var(r^*)\geq Var(r^*_{(1)}) + Var(r^*_{(2)})$
proofWe consider two time series $X$ and $Y$, where: $(X^{(1)},Y^{(1)}) $ and $(X^{(2)},Y^{(2)})$ are pairs of signals band-limited to disjoint frequency ranges $\Omega_1$ and $\Omega_2$, with $\Omega_1\cap \Omega_2= \emptyset$ and each pair admit a variable-lag Granger representation via alignment path $P^{(1)},P^{(2)}$ minimizing residuals. Since $r^*_{(i)}$ be the residual from fitting $Y^{(i)} \sim VL-G(X^{(i)};P^{(i)})$ and $r^*$ be the residual from fitting a single VL-Granger model to $(X,Y)$, with alignment $P$. So, we can show $$Y_t = f_1(X_{t-\delta^{(1)}(t)})+f_2(X_{t-\delta^{(2)}(t)})+\varepsilon_t$$ where $\delta^{(i)}(t)$ is a nonlinear time lag, $f_i$ is the optimal regressor under alignment $\delta^{(i)}$, and $\varepsilon_t$ is i.i.d. noise and let $\hat{Y}_t^{(i)}=f_i(X_{t-\delta^{(i)}(t)})$ We fit a path $\delta(t)$ to model: $\hat{Y}_t = g(X_{t-\delta(t)})$. Then, $X_{t-\delta (t)}= X_{t-\delta (t)}^{(1)}+X_{t-\delta (t)}^{(2)}$. So, $X_{t-\delta (t)}^{(1)}+X_{t-\delta (t)}^{(2)} \geq X_{t-\delta^{(1)} (t)}^{(1)}+X_{t-\delta^{(2)} (t)}^{(2)}$ this means $g$ cannot access both true lags. Hence, $r^* = Y_t -g(X_{t-\delta(t)})=r^*_{(1)}+r^*_{(2)}+\eta_t$. Then, $$Var(r^*)\geq Var(r^*_{(1)}) + Var(r^*_{(2)})$$ Note that: For the not perfect fitting in $(X,Y)$ case, $\eta_t$ keeps positive, since there must be a information missing. It implies $$Var(r^*)> Var(r^*_{(1)}) + Var(r^*_{(2)})$$.

With the base case proved in Proposition 2.2, Proposition 2.3 generalizes the result to $n-$band separation, which we state without proof via induction.

proposition$Var(r^*)\geq \sum_{i=1}^n Var(r^*_{(i)}) $

In conclusion, we can say that by dropping a stationary and a VAR assumption, the VL-Granger causality test can be applied on the multi-decomposed bands of the time series. In the case of not perfect fitting in $(X,Y)$ ($\eta_t>0$), $Var(r^*)> \sum_{i=1}^n Var(r^*_{(i)}) $, which implies MB-VLGC performs better than VLGC.

Methods

Frequency-Band VL-Granger Framework Architecture

Fig (ref) shows the high-level overview of our framework. Our key idea lies in the systematic application of VLGC to frequency-decomposed signals, enabling frequency-specific causality detection that had not been done with previous methods constrained by filter invariance. This allows detection of causal relationships with different temporal delays across frequency bands and frequency-specific causality masked in broadband analysis—capabilities essential for understanding multi-timescale biological and physical processes.

Framework Pipeline

There are three steps in the pipeline and the details of each step are below.

Stage 1: Frequency Banding. Input time series are decomposed using bandpass filters into frequency-specific components for each target band. This stage needs to preserves temporal relationships while achieving clean frequency separation essential for reliable causality detection.

Stage 2: Causal Inference. VLGC analysis is applied independently to each frequency band, leveraging the theoretical capability of VLGC to operate on filtered signals without losing causal information. Each band yields detection statistics, lag estimates, and confidence measures.

Stage 3: Result Integration. Band-specific results statistic are systematically combined using established method and cross-band consistency assessment, producing both overall causality decisions and frequency-specific insights.

Band decomposition

As temporal preservation is needed for Granger Causality because it relies on using past observation to predict future values. Zero-phase filtering was implemented using the filtfilt function, which performs forward and backward filtering to eliminate phase distortion gustafsson1996determining and preserve its temporal properties.

We employed 4th-order Butterworth bandpass filters due to their maximally flat frequency response in the passband, ensuring minimal amplitude distortion of signal components oppenheim1999discrete. Studies examining neural connectivity consistently employ Butterworth filters for frequency band isolation due to their minimal artifacts and their maximally flat frequency response in the passband bastos2015tutorial, cohen2014analyzing.

Per-Band Causality Analysis

Our frequency-band method builds upon VLGC 10.1145/3441452, which extends traditional Granger causality to handle time-varying delays between cause and effect. While traditional Granger causality assumes fixed temporal lags, VLGC allows causal relationships to exhibit variable delays that change dynamically over time, making it more suitable for real-world phenomena where causal timing is not constant.

Dynamic Time Warping Alignment

VLGC addresses variable delays by using Dynamic Time Warping (DTW) to find optimal temporal alignments between the cause time series $X$ and effect time series $Y$. DTW identifies the warping path $P^* = (\Delta_1, \Delta_2, ..., \Delta_T)$ that minimizes the cumulative distance between aligned elements: $$P^* = \arg\min_P \sum_{t} D(X(t-\Delta_t), Y(t))$$ where $\Delta_t$ represents the time delay at step $t$, and $D(\cdot,\cdot)$ is the distance function between aligned points.

Variable-Lag Regression Framework

Using the optimal warping path, VLGC constructs time-aligned predictors for regression analysis. The variable-lag causality test compares three nested models:

Null Model ($H_0$): $Y(t) = \sum_{i=1}^{\delta_{max}} a_i Y(t-i) + \epsilon_Y(t)$ \\ \\ Fixed-Lag Model ($H_1$): $Y(t) = \sum_{i=1}^{\delta_{max}} a_i Y(t-i) + \sum_{i=1}^{\delta_{max}} b_i X(t-i) + \epsilon_{YX}(t)$ \\ \\ Variable-Lag Model ($H_2$): $Y(t) = \sum_{i=1}^{\delta_{max}} a_i Y(t-i) + \sum_{i=1}^{\delta_{max}} c_i X^*(t-i) + \epsilon_{VL}(t)$ \\ \\ where $X^*(t-i) = X(t-i+1-\Delta_{t-i+1})$ represents the DTW-aligned version of $X$, and $\delta_{max}$ is the maximum lag considered.

Hybrid Lag Selection Strategy

VLGC employs a hybrid approach combining cross-correlation and DTW for optimal lag selection at each time point. The process begins with cross-correlation analysis to identify the globally optimal delay:

$$opt_{delay} = \arg\max_{\tau} |CCF(\tau)| $$ $$= \arg\max_{\tau} \left|\frac{\sum_{t} (X(t-\tau) - \bar{X})(Y(t) - \bar{Y})}{\sqrt{\sum_{t}(X(t) - \bar{X})^2 \sum_{t}(Y(t) - \bar{Y})^2}}\right|$$

where $CCF(\tau)$ is the cross-correlation function at lag $\tau$, and $\bar{X}$, $\bar{Y}$ are sample means.

At each time point $t$, the method selects the lag that minimizes prediction error between candidate alignments: $chosen_{lag}(t) = \arg\min_{\tau \in \{\tau_{CC}, \tau_{DTW}(t)\}} |Y(t) - X(t - \tau)|$

where $\tau_{CC}$ is the global cross-correlation lag and $\tau_{DTW}(t)$ is the DTW-suggested lag at time $t$. This hybrid strategy combines the global optimality of cross-correlation with the local adaptivity of DTW, yielding time-specific alignments that maximize the causal signal strength while maintaining temporal coherence.

Statistical Testing Framework

VLGC employs dual criteria for causality detection:

The method concludes $X$ causes $Y$ if either criterion is satisfied: $p_{F-test} \leq \alpha$ OR $\gamma \geq \gamma_{threshold}$, providing robustness against different signal characteristics and noise conditions.

Result integration

P-value Combination.

Valid p-values from individual frequency bands are combined using established meta-analysis methods. We employ Fisher's combined probability test as the primary approach: $$\chi^2 = -2\sum_{i=1}^{k} \ln(p_i)$$ where $k$ is the number of valid frequency bands. Alternative combination methods (Stouffer's method, Bonferroni correction) are also available.

Experimental Evaluation

Experimental Setup

We conducted comprehensive experiments to evaluate our multi-band VL-Granger causality method across two main objectives. First, we tested the method's capability to handle various types of causal relationships. Second, we assessed its performance in identifying causal relationships in real-world datasets. \footnote{All experiments ran on MacBook Air M2,2022 Memory 16 GB SSD 256 GB}

We compared our proposed method against five established causality detection approaches:

itemize• Variable-lag Granger Causality (VL-GC) 10.1145/3441452: The foundational variable-lag method that our approach extends • Granger Causality (GC) granger1969investigating: Traditional fixed-lag Granger causality as implemented in standard econometric packages • Variable-lag Transfer Entropy (VL-TE) 10.1145/3441452: Non-linear extension of transfer entropy with variable-lag capability • Transfer Entropy (TE) BEHRENDT2019100265: Standard transfer entropy with fixed-lag assumptions • PCMCI+ runge2020discovering: State-of-the-art conditional independence-based causality detection method • Granger-Geweke (GG) farne2022bootstrap: Spectral decomposition of Granger causality that measures causal strength in the frequency domain using R-CRAN Package \textit{grangers}.

For all experiments, we set the significance level $\alpha = 0.01$ and the BIC difference ratio threshold $\gamma = 0.6$ by default, unless explicitly stated otherwise.

Datasets

Synthetic Datasets

We generated a comprehensive synthetic dataset consisting of 240 time series files to systematically evaluate causality detection performance across different scenarios. The dataset is divided into two main categories: datasets with ground truth causality (120 files) and datasets without causality (120 files) for testing false positive rates.

\paragraph{False Positive Test Datasets (120 files):} Independent time series generated from standard normal distributions with no causal relationships (Random Noise).

\paragraph{True Positive Test Datasets (120 files):} The causality datasets are further divided into four distinct types:

Basic Causation (30 files): Simple linear causality where the effect time series is generated from the cause time series with a fixed lag of 20 samples and random coupling coefficient.

Variable-lag Causation (30 files): The effect follows the cause with discrete lag periods that change over time. The lag switches between three values (12, 16, and 20 samples) with 2-4 regime changes per time series.

Broadband Causation (30 files): Both cause and effect contain rich frequency content spanning 1-50 Hz, with the effect generated by lagging the broadband cause signal by 7 samples. This tests performance on frequency-rich signals where traditional methods may struggle.

Multi-frequency Causation (30 files): Complex scenarios where different frequency components (10, 40, and 80 Hz) exhibit different causal lags (15, 8, and 4 samples respectively). This directly tests the method's ability to detect frequency-specific causal relationships.

All synthetic time series were generated with lengths between 500-2000 samples at a sampling rate of 250 Hz. Random seeds were set for reproducibility, and coupling coefficients were drawn from normal distributions to ensure realistic signal-to-noise ratios. The variable-lag datasets use discrete lag periods rather than continuous variation to create detectable patterns that reflect realistic biological and physical processes. The code for data generation is available in the code and datasets link.

Real-world Datasets

We evaluated our method on four established real-world datasets from diverse domains to assess its practical applicability and validate frequency-specific causality detection capabilities. The $X$ represents case and $Y$ represents effect in the setting.

\paragraph{Old Faithful Geyser:} This classic dataset contains eruption duration ($X$) and inter-eruption intervals $Y$ from Old Faithful geyser azzalini1990look. We tested causality between consecutive eruption characteristics, where previous eruption duration may influence the next inter-eruption interval. The dataset contains 298 observations and serves as a benchmark for temporal causality methods.

\paragraph{Chicken and Egg Prices:} Economic time series data examining the price relationship between egg $X$ and chicken $Y$ markets. This dataset zeileis2002diagnostic tests our method's ability to detect economic causality relationships that may operate at different frequency scales, from short-term market fluctuations to longer-term supply-demand dynamics. The time series length is 54 time steps.

\paragraph{Gas Furnace:} Industrial process data measuring gas consumption rates $X$ and corresponding CO$_2$ output $Y$ box2015time. This controlled system provides a clear causal relationship where gas input drives CO$_2$ production, with the dataset containing 296 time steps. The known causal direction makes this ideal for validation.

\paragraph{EEG Motor Imagery:} Neurophysiological data from the EEG Motor Movement/Imagery Dataset schalk2004bci2000 available through PhysioNet. We analyzed causality between electrodes FC3 and FC5, which are positioned over motor cortex regions. These electrodes are expected to show coordinated activity during motor tasks, with potential frequency-specific interactions in different neural rhythms (alpha, beta, gamma bands). The high sampling rate (250 Hz) and rich frequency content make this dataset particularly suitable for testing frequency-band causality detection.

These datasets span multiple domains (geophysical, economic, industrial, and neurophysiological) and provide diverse signal characteristics including different sampling rates, noise levels, and frequency content. The neurophysiological data is especially relevant for demonstrating the practical value of frequency-specific causality analysis, as different neural rhythms are known to carry distinct functional information.

Results

We outline the result of testing our approach against five established methods on both synthetic and real-world datasets.

Synthetic Dataset Performance

table[table omitted — 1,154 chars of source]

Our method achieved the highest overall F1-score of 0.810. Most notably, our method excelled on multi-frequency datasets with an accuracy of 0.933, significantly outperforming traditional methods that struggled with frequency-specific causal relationships. This demonstrates the core strength of our approach: the ability to detect causality that operates differently across frequency bands. For the case of positive class (X causes Y), PCMCI+ performed the best and it performed slightly better than our method (MB-VL). However, it performed worse than our method in the random noise case (X does not cause Y). The VLGC and GC performed the best in the random noise case but they were unable to deal with many types of lag case especially the Multifrequency lag case. The VLTE performed better than TE and our method in most of the cases but it performed dramaically poor compare to our method in the Multifrequency lag case. Lastly, GG had the worst performance.

Impact of Frequency Band Configuration

A critical aspect of our method is the selection of frequency bands, which directly impacts performance depending on the underlying signal characteristics. Table (ref) illustrates how different band configuration strategies affect performance.

table[table omitted — 443 chars of source]

The two-band configuration (1-80 Hz and 81-120 Hz) provided the optimal balance across all dataset types, achieving the highest overall F1-score while maintaining excellent performance on multi-frequency datasets. This configuration represents a principled frequency division that separates low-frequency oscillatory dynamics from higher-frequency transient processes.

Interestingly, the EEG-specific six-band configuration achieved perfect performance (1.000) on multi-frequency datasets but lower overall performance (0.618) due to over-segmentation of simpler causal relationships. This illustrates an important principle: Optimal band selection depends on the expected characteristics of the causal relationships and available domain knowledge.

For applications where signal characteristics are unknown, our two-band configuration provides robust performance. However, when domain-specific knowledge is available—such as in EEG analysis where neural oscillations operate in well-defined frequency ranges—informed band selection allows users to investigate frequency-specific causal relationships that correspond to their research questions and domain understanding, revealing underlying mechanisms that would remain hidden in typical causal analysis.

Frequency-Specific Lag Detection

Our method provides insights into the temporal dynamics within different frequency bands beyond causality detection. Table (ref) demonstrates our method's ability to accurately detect causal lags across different neural frequency bands from the multi-frequency causation dataset.

table[table omitted — 519 chars of source]

The results show that our method successfully detects frequency-specific temporal delays with high accuracy. High gamma frequencies (30-100 Hz) demonstrated the most precise lag detection, closely matching the true lag of 4 samples. Alpha band detection showed more variability, which is consistent with the known properties of alpha oscillations in neural systems. This frequency-specific lag information provides valuable insights that would be lost in traditional broadband analysis.

Real-world Dataset Validation

figure*[figure* omitted — 432 chars of source]

To validate practical applicability, we tested our method on four established real-world datasets from diverse domains. Table (ref) presents the causality detection results, where 1 indicates successful causality detection and 0 indicates no detected causality.

table[table omitted — 591 chars of source]

Our method achieved perfect causality detection across all four real-world datasets, demonstrating robust performance in practical applications. The EEG motor imagery dataset particularly showcases the value of frequency-specific analysis, where our method successfully detected causality between FC3 and FC5 electrodes while the traditional Granger causality failed. This result aligns with neuro-scientific understanding that motor cortex regions exhibit coordinated activity across different neural frequency bands.

In Fig. (ref) (the right bottom), only the gamma band has the bidirectional VL-Granger causation which is consistent with the ground truth while there were no causal relation in other bands. This implies that the gamma band might be the main contributor for the causal relations between FC3 and FC5 electrodes.

The diverse nature of these datasets—spanning geophysical (Old Faithful geyser), economic (chicken and egg prices), industrial (gas furnace), and neuro-physiological (EEG) domains—demonstrates the broad applicability of our frequency-band approach. Notably, PCMCI+ failed on the economic dataset, highlighting scenarios where conditional independence assumptions may be violated but frequency-specific causality relationships still exist.

Conclusion

We formalized Multi-Band Variable-Lag Granger Causality (MB-VLGC) and proposed a novel framework that generalizes traditional variable-lag Granger causality (VLGC) by explicitly modeling frequency-dependent causal delays. We provided a formal definition of MB-VLGC, demonstrated its theoretical soundness, and proposed an efficient inference pipeline.

According to the results, our Multi-Band VL-Granger causality method addressed fundamental limitations in existing causality detection approaches by enabling frequency-specific analysis. The method achieved the better overall performance (F1 = 0.810) than others (Table (ref)) while providing the flexibility to adapt to domain-specific requirements through informed band selection. The key finding is that optimal performance requires matching the frequency band configuration to the expected characteristics of causal relationships, with our two-band default configuration providing robust performance when domain knowledge is limited. Code and datasets are publicly available in Code and datasets section.

links\link{Code and datasets}{https://anonymous.4open.science/r/mbvlgranger-ED89/README.md}
thebibliography{21} \bibitem[{Amornbunchornvej, Zheleva, and Berger-Wolf(2021)}]{10.1145/3441452} Amornbunchornvej, C.; Zheleva, E.; and Berger-Wolf, T. 2021. \newblock Variable-Lag Granger Causality and Transfer Entropy for Time Series Analysis. \newblock ACM Trans. Knowl. Discov. Data, 15(4). \bibitem[{Azzalini and Bowman(1990)}]{azzalini1990look} Azzalini, A.; and Bowman, A. W. 1990. \newblock A look at some data on the Old Faithful geyser. \newblock Journal of the Royal Statistical Society: Series C (Applied Statistics), 39(3): 357--365. \bibitem[{Bastos and Schoffelen(2015)}]{bastos2015tutorial} Bastos, A. M.; and Schoffelen, J.-M. 2015. \newblock A tutorial review of functional connectivity analysis methods and their interpretational pitfalls. \newblock Frontiers in Systems Neuroscience, 9: 175. \bibitem[{Behrendt et al.(2019)Behrendt, Dimpfl, Peter, and Zimmermann}]{BEHRENDT2019100265} Behrendt, S.; Dimpfl, T.; Peter, F. J.; and Zimmermann, D. J. 2019. \newblock RTransferEntropy — Quantifying information flow between different time series using effective transfer entropy. \newblock SoftwareX, 10: 100265. \bibitem[{Box et al.(2015)Box, Jenkins, Reinsel, and Ljung}]{box2015time} Box, G. E.; Jenkins, G. M.; Reinsel, G. C.; and Ljung, G. M. 2015. \newblock Time series analysis: forecasting and control. \newblock John Wiley & Sons. \bibitem[{Canolty and Knight(2010)}]{canolty2010functional} Canolty, R. T.; and Knight, R. T. 2010. \newblock The functional role of cross-frequency coupling. \newblock Trends in cognitive sciences, 14(11): 506--515. \bibitem[{Cohen(2014)}]{cohen2014analyzing} Cohen, M. X. 2014. \newblock \emph{Analyzing neural time series data: theory and practice}. \newblock MIT press. \bibitem[{Farn{\`e} and Montanari(2022)}]{farne2022bootstrap} Farn{\`e}, M.; and Montanari, A. 2022. \newblock A bootstrap method to test Granger-causality in the frequency domain. \newblock \emph{Computational Economics}, 59(3): 935--966. \bibitem[{Geweke(1982)}]{Geweke1982} Geweke, J. 1982. \newblock Measurement of linear dependence and feedback between multiple time series. \newblock \emph{Journal of the American Statistical Association}, 77(378): 304--313. \bibitem[{Granger(1969)}]{granger1969investigating} Granger, C. W. 1969. \newblock Investigating causal relations by econometric models and cross-spectral methods. \newblock \emph{Econometrica: journal of the Econometric Society}, 424--438. \bibitem[{Granger and Newbold(1974)}]{granger1974spurious} Granger, C. W. J.; and Newbold, P. 1974. \newblock Spurious regressions in econometrics. \newblock \emph{Journal of Econometrics}, 2(2): 111--120. \bibitem[{Gustafsson(1996)}]{gustafsson1996determining} Gustafsson, F. 1996. \newblock Determining the Initial States in Forward-Backward Filtering. \newblock \emph{IEEE Transactions on Signal Processing}, 44(4): 988--992. \bibitem[{Hosoya(1991)}]{hosoya1991decomposition} Hosoya, Y. 1991. \newblock The decomposition and measurement of the interdependency between second-order stationary processes. \newblock \emph{Probability theory and related fields}, 88(4): 429--444. \bibitem[{Michalareas et al.(2016)Michalareas, Vezoli, Van Pelt, Schoffelen, Kennedy, and Fries}]{michalareas2016alpha} Michalareas, G.; Vezoli, J.; Van Pelt, S.; Schoffelen, J.-M.; Kennedy, H.; and Fries, P. 2016. \newblock Alpha-beta and gamma rhythms subserve feedback and feedforward influences among human visual cortical areas. \newblock \emph{Neuron}, 89(2): 384--397. \bibitem[{Oppenheim(1999)}]{oppenheim1999discrete} Oppenheim, A. V. 1999. \newblock \emph{Discrete-time signal processing}. \newblock Pearson Education India. \bibitem[{Runge(2020)}]{runge2020discovering} Runge, J. 2020. \newblock Discovering contemporaneous and lagged causal relations in autocorrelated nonlinear time series datasets. \newblock In \emph{Conference on uncertainty in artificial intelligence}, 1388--1397. Pmlr. \bibitem[{Sakoe and Chiba(1978)}]{1163055} Sakoe, H.; and Chiba, S. 1978. \newblock Dynamic programming algorithm optimization for spoken word recognition. \newblock \emph{IEEE Transactions on Acoustics, Speech, and Signal Processing}, 26(1): 43--49. \bibitem[{Schalk et al.(2004)Schalk, McFarland, Hinterberger, Birbaumer, and Wolpaw}]{schalk2004bci2000} Schalk, G.; McFarland, D. J.; Hinterberger, T.; Birbaumer, N.; and Wolpaw, J. R. 2004. \newblock BCI2000: a general-purpose brain-computer interface (BCI) system. \newblock \emph{IEEE Transactions on biomedical engineering}, 51(6): 1034--1043. \bibitem[{Schreiber(2000)}]{Schreiber2000} Schreiber, T. 2000. \newblock Measuring information transfer. \newblock \emph{Physical Review Letters}, 85(2): 461--464. \bibitem[{Varian(2016)}]{varian2016causal} Varian, H. R. 2016. \newblock Causal inference in economics and marketing. \newblock \emph{Proceedings of the National Academy of Sciences}, 113(27): 7310--7315. \bibitem[{Zeileis and Hothorn(2002)}]{zeileis2002diagnostic} Zeileis, A.; and Hothorn, T. 2002. \newblock Diagnostic Checking in Regression Relationships. \newblock \emph{R News}, 2(3): 7--10.