Fact-checked by Grok 2 weeks ago

Circular mean

The circular mean, also known as the angular mean or mean direction, is a fundamental measure of in directional statistics, designed for data that are periodic and measured on a , such as , clock times, or orientations. It addresses the limitations of the , which fails for circular data because wrap around (e.g., 359° and 1° should near 0°, not 180°). For a sample of \theta_1, \dots, \theta_n in radians, the circular mean \mu is defined as \mu = \atantwo\left( \frac{1}{n} \sum_{i=1}^n \sin \theta_i, \frac{1}{n} \sum_{i=1}^n \cos \theta_i \right), or equivalently, the argument of the \arg\left( \frac{1}{n} \sum_{i=1}^n e^{i \theta_i} \right), yielding a value in [0, 2\pi). This computation aligns with the resultant vector's direction on the unit , ensuring rotational invariance and respect for the data's . Within the broader framework of directional statistics, the circular mean quantifies location for unimodal circular distributions, such as the , which serves as the "circular normal" and is parameterized by a mean direction \mu and concentration \kappa. Its properties include asymptotic unbiasedness under the von Mises model and utility in hypothesis testing, like the Rayleigh test for uniformity, where the mean resultant length R = \left| \frac{1}{n} \sum_{i=1}^n e^{i \theta_i} \right| measures dispersion around the (with R = 1 indicating no spread and R = 0 full uniformity). The concept extends to higher dimensions as the von Mises-Fisher mean on spheres, but the circular case remains central for 2D applications. Foundational developments trace to works like Mardia's Statistics of Directional Data (1972) and the comprehensive Directional Statistics by Mardia and Jupp (2000), which formalize these methods. The circular mean finds widespread use across disciplines involving oriented or cyclic data. In , it analyzes directions, such as paths or orientations, enabling tests for preferred headings amid environmental cues. In , it computes average wind or current directions from angular measurements, as seen in time series models for airport wind data. Geological applications include estimating mean orientations of rock fabrics or fault strikes, while employs it for phases. These implementations are supported in statistical software like and R's circular package, facilitating robust inference.

Definition

Unit vector approach

Circular data consists of angles \theta_i, where each \theta_i lies in the interval [0, 2\pi), representing directions or orientations that wrap around a circle. To compute the circular mean, these angles are represented geometrically as unit vectors in the plane, with each observation corresponding to the point (\cos \theta_i, \sin \theta_i) on the unit circle. This vector representation preserves the cyclic nature of the data, avoiding distortions that arise from treating angles as linear values. The circular mean \mu is derived from the average of these unit vectors. Let n be the number of observations; the mean vector components are given by \bar{C} = \frac{1}{n} \sum_{i=1}^n \cos \theta_i and \bar{S} = \frac{1}{n} \sum_{i=1}^n \sin \theta_i. The length of this resultant vector is R = \sqrt{\bar{C}^2 + \bar{S}^2}, which measures the concentration of the data (with R = 1 indicating perfect alignment and R = 0 indicating uniform dispersion). The circular mean direction is then \mu = \atan2(\bar{S}, \bar{C}), where \atan2 ensures the correct quadrant by considering the signs of \bar{S} and \bar{C}. This approach follows from basic principles of vector addition, where the sum of unit vectors yields a whose direction defines the . Geometrically, the unit vector approach interprets the circular mean as the angle of the resultant vector obtained by summing the individual unit vectors head-to-tail on the unit circle. This method naturally handles the wrap-around property of circular data, preventing ambiguities such as those encountered when averaging directly (e.g., the mean of 1° and 359° is 0° or 360°, not 180°). The resultant length R provides an intuitive measure of how closely the angles cluster around \mu, with shorter vectors indicating greater .

Complex exponential representation

The circular mean can be equivalently represented using complex exponentials, where each angle \theta_j is mapped to a point on the unit circle in the via e^{i \theta_j} = \cos \theta_j + i \sin \theta_j. For a sample of n angles \theta_1, \dots, \theta_n, the circular mean \mu is given by \mu = \arg\left( \frac{1}{n} \sum_{j=1}^n e^{i \theta_j} \right), where \arg denotes the argument (principal value) of the , yielding an angle in (-\pi, \pi]. This formulation treats the angles as phasors, with the mean direction emerging as the phase of their sum normalized by n. This complex exponential approach derives its equivalence to the unit vector method through , e^{i \theta} = \cos \theta + i \sin \theta. Substituting yields \sum_{j=1}^n e^{i \theta_j} = \sum_{j=1}^n \cos \theta_j + i \sum_{j=1}^n \sin \theta_j, so the argument is \arg\left( \sum \cos \theta_j + i \sum \sin \theta_j \right) = \atantwo\left( \sum \sin \theta_j, \sum \cos \theta_j \right), matching the direction of the from the real and imaginary components. For the population mean, the analogous expression is \mu = \arg\left( \mathbb{E}[e^{i \theta}] \right), where \mathbb{E}[e^{i \theta}] = \rho e^{i \mu} and \rho is the population mean length, a measure of concentration ranging from 0 (uniform ) to 1 (no ). The primary advantages of this lie in its algebraic compactness: it consolidates the into a single complex operation, obviating the need for separate evaluations, and directly provides the \left| \frac{1}{n} \sum e^{i \theta_j} \right| as the sample mean resultant length \bar{R}, which quantifies data concentration without additional . This facilitates analytical tractability in circular statistical models, such as the , where \mathbb{E}[e^{i \theta}] = I_1(\kappa)/I_0(\kappa) \cdot e^{i \mu} and I_r are modified of the first kind.

Properties

Mathematical characteristics

The circular mean satisfies several axiomatic properties that establish it as a valid measure of for directional data. Rotational invariance holds, meaning that adding a constant c to all observations shifts the circular mean by exactly c $2\pi. is also satisfied, such that the circular mean of a set of circular means coincides with the overall circular mean. Additionally, the identity ensures that if all angles are identical, the circular mean equals that common . A key characteristic is the mean resultant length \rho, defined for a sample of n angles \theta_j as \rho = \left| \frac{1}{n} \sum_{j=1}^n e^{i \theta_j} \right|, which quantifies the concentration of the data around the mean direction. This scalar ranges from 0, indicating uniform dispersion with no preferred direction, to 1, signifying perfect alignment of all angles. For distributions like the von Mises, \rho = A(\kappa) = I_1(\kappa)/I_0(\kappa), where I_\nu denotes the modified of the first kind of order \nu, and \kappa is the concentration parameter. In finite samples, the sample mean resultant length R/n (where R = \left| \sum e^{i \theta_j} \right|) is biased toward 0 as an estimator of the population \rho. This bias arises due to the nonlinear nature of the modulus operation and can be corrected using factors derived from ; for instance, the maximum likelihood estimator for the concentration \kappa in the von Mises case is \hat{\kappa} = A^{-1}(R/n), which is biased (low) especially in small samples and requires correction. The circular mean is unique for any non-uniform distribution but undefined for the uniform distribution, where the resultant sums to zero (\rho = 0) and no preferred direction exists.

Comparison with linear mean

The fails for circular data because it treats angles as linear values on an unbounded scale, ignoring the periodic nature of the circle where 0° and 360° are equivalent. For instance, the of 1° and 359° is 180°, which points in the opposite direction from the intuitive average near 0°, leading to paradoxes in directional interpretation. Similarly, for angles 10°, 30°, and 350°, the yields 130° (southeast), whereas the data cluster near north (0°). When data concentration is high, indicated by the mean resultant length ρ close to 1, the circular mean approximates the of the angles, as the (modeling concentrated circular data) behaves like a on the line. This approximation holds because small angular deviations allow linear averaging without significant wrap-around effects, but it breaks down for dispersed data. Examples of divergence include uniform distributions, where the circular mean is undefined (ρ ≈ 0, no preferred direction), but the arithmetic mean arbitrarily selects a midpoint like 180° for evenly spaced points. In contrast, for clustered data (e.g., angles tightly grouped around 45°), both means align closely, but the arithmetic mean distorts results for bimodal or wrapping clusters, such as directions split across 0°/360°. Statistically, the circular mean minimizes the angular squared error on the circle by maximizing the resultant vector length, thereby minimizing circular variance (1 - ρ), unlike the arithmetic mean, which minimizes Euclidean squared error on the line and ignores toroidal geometry. This property ensures the circular mean provides the maximum-likelihood estimate under common models like the von Mises distribution for directional data.

Estimation and Computation

Maximum likelihood estimation

The maximum likelihood estimation (MLE) of the circular mean assumes that the observed angles \theta_1, \dots, \theta_n are independent and identically distributed from a with unknown mean direction \mu and concentration parameter \kappa > 0, which has probability density function f(\theta; \mu, \kappa) = \frac{1}{2\pi I_0(\kappa)} \exp\{\kappa \cos(\theta - \mu)\} for \theta \in [0, 2\pi), where I_0(\kappa) is the modified Bessel function of the first kind of order zero. Under this model, the MLE for \mu is \hat{\mu} = \arg\left( \sum_{j=1}^n e^{i \theta_j} \right), which coincides with the sample circular mean defined as the argument of the resultant vector sum. This estimator maximizes the log-likelihood l(\mu, \kappa) = -n \log(2\pi I_0(\kappa)) + \kappa \sum_{j=1}^n \cos(\theta_j - \mu) with respect to \mu. The MLE \hat{\mu} is consistent and asymptotically efficient as n \to \infty, with asymptotic normality \sqrt{n} (\hat{\mu} - \mu) \xrightarrow{d} \mathcal{N}\left(0, \frac{1}{\kappa A(\kappa)}\right), where A(\kappa) = I_1(\kappa)/I_0(\kappa) is the population mean resultant length; for large \kappa, this variance approximates $1/(n \kappa). Joint estimation of \kappa proceeds by first computing the sample resultant length R = \left| \sum_{j=1}^n e^{i \theta_j} \right| / n, then solving \hat{\kappa} = A^{-1}(R) iteratively, as A(\cdot) lacks a closed-form inverse and requires numerical methods such as Newton-Raphson or lookup tables based on the ratio of . For small samples, the MLE \hat{\kappa} exhibits approximately E[\hat{\kappa} - \kappa] \approx -3\kappa / (5n) for small \kappa, which can be corrected using higher-order expansions such as the adjusted \hat{\kappa}^* = \hat{\kappa} \left[ 1 + \frac{A'(\hat{\kappa})}{\ n \hat{\kappa} A(\hat{\kappa})} \right], where A'(\kappa) is the of A(\kappa).

Numerical methods

The direct computation of the circular mean for unimodal data involves representing each angle as a , summing the components, and then applying the two-argument arctangent function to the summed coordinates. Specifically, for angles \theta_i (in radians), compute the sums C = \sum \cos \theta_i and S = \sum \sin \theta_i, then the mean is \bar{\theta} = \atantwo(S, C), where \atantwo handles the correct . Alternatively, using complex exponentials, the mean can be obtained as \bar{\theta} = \arg\left( \sum e^{i \theta_i} \right), which is equivalent for unimodal distributions. This approach corresponds to the maximum likelihood estimator under the for unimodal data. This direct method operates in O(n) , where n is the number of observations, as it requires a single pass to sum the vector components. For large datasets, efficiency can be improved through parallel summation of the components, leveraging vectorized operations in numerical libraries. Edge cases must be handled explicitly: if all angles are identical, the equals that with resultant length \rho = 1; if the data are uniformly distributed, the is , typically reported alongside \rho = 0 to indicate . For multimodal data, where multiple clusters exist on the circle, direct computation may yield misleading results; instead, Bayesian mixture models or clustering algorithms can identify components before computing a principal from the dominant mode. Bayesian approaches, such as mixtures of von Mises distributions, infer the number of modes and their parameters via sampling. Clustering methods like circular k-means adapt the standard k-means by using angular distances to form circular-invariant clusters, from which the of the largest cluster serves as the principal direction. Software implementations facilitate these computations. In R, the circular package provides mean.circular(), which sums unit vectors and applies atan2 internally, supporting both radians and degrees. In Python, scipy.stats.circmean from SciPy computes the mean via vector summation and math.atan2, with options for periodicity handling. In MATLAB, the CircStat toolbox offers circ_mean, which performs the unit vector average and returns both the mean and resultant length, optimized for directional data analysis.

Examples

Simple numerical example

Consider a simple of three measured in degrees: 0°, 10°, and 350°. These values illustrate the wrap-around nature of circular , as 350° is equivalent to -10° and lies close to 0° on the circle. To compute the circular mean using the unit vector approach, first convert the to radians: 0 rad, approximately 0.175 rad, and approximately 6.109 rad. Next, calculate the of the cosines and sines: C = \sum \cos \theta_i = \cos(0) + \cos(0.175) + \cos(6.109) \approx 1 + 0.985 + 0.985 = 2.970 S = \sum \sin \theta_i = \sin(0) + \sin(0.175) + \sin(6.109) \approx 0 + 0.174 - 0.174 = 0 The resultant vector length is R = \sqrt{C^2 + S^2} \approx 2.970, and the mean resultant length (concentration measure) is \bar{R} = R / n \approx 2.970 / 3 \approx 0.990, where n = 3 is the sample size. The circular mean angle \mu is given by \atantwo(S, C) \approx 0 radians, or 0° (noting that the two-argument arctangent function ensures the correct quadrant). This result demonstrates high concentration (\bar{R} \approx 0.99), as the angles cluster tightly around 0° despite the apparent spread when treated linearly—the arithmetic mean would be (0 + 10 + 350)/3 = 120°, which misrepresents the central tendency. In contrast, the circular mean correctly identifies the direction near the data cluster, accounting for the modular arithmetic of the circle. For visualization, plotting the unit vectors at these angles on a unit circle reveals their near-alignment, with the resultant vector pointing directly along the positive x-axis at 0°.

Practical applications

The circular mean is widely used in to determine the prevailing from multiple bearings, such as hourly observations, by converting directions to unit vectors, averaging their components, and computing the resultant angle, which avoids distortions from linear averaging across the 360-degree cycle. For instance, in wind energy assessments, joint models of and direction employ circular statistics to forecast probabilistic distributions, enabling accurate site evaluations for turbine placement. This approach is essential for regulatory modeling, where mean wind vectors from data inform simulations for air quality predictions. In , particularly in , the circular quantifies animal orientation behaviors, such as the average heading of during tracked via radio or funnel tests. Researchers apply it to analyze head cells in the brains of migratory , calculating the vector length and to assess navigational cues like , with significant orientation often yielding resultant lengths above 0.5. For example, in studies of Eurasian reed warblers, circular means from Emlen reveal preferred routes, distinguishing innate from learned behaviors with vector lengths indicating clustering around the . Astronomy utilizes the circular mean to average positional of galaxies, determining overall orientations in large surveys to investigate with cosmic structures. In polarization studies, it extracts rotation measures from data by maximizing likelihood over angular distributions, crucial for mapping magnetic fields in radio galaxies. Galaxy alignment analyses further employ circular statistics to test uniformity of position , revealing correlations with large-scale filaments that inform models. In engineering, particularly , the circular mean processes phase angles in circularly symmetric , such as those from uniform circular arrays in or wireless communications, to estimate without phase wrapping errors. For non-destructive testing, it measures shifts in electromagnetic signals using circular variance to quantify signal , with applications in material flaw detection. measurements also leverage circular means for , applying them alongside linear statistics for magnitude to evaluate beam patterns accurately. Historically, the circular mean found early use in 19th-century for analyzing directional data, predating formal directional statistics. Ronald A. Fisher advanced the field in the through foundational work on spherical , motivated by astronomical and geological problems, which laid groundwork for modern applications despite his key publication appearing in 1953.

Generalizations and Extensions

Weighted circular mean

The weighted circular mean generalizes the circular mean to incorporate varying reliabilities among observations via non-negative weights w_j > 0 associated with angles \theta_j, j = 1, \dots, n. The weighted resultant vector is given by \bar{R} e^{i \mu} = \frac{\sum_{j=1}^n w_j e^{i \theta_j}}{\sum_{j=1}^n w_j}, where \mu is the weighted mean direction and \bar{R} is the weighted mean resultant length, a measure of angular concentration ranging from 0 (uniform dispersion) to 1 (perfect alignment). When all weights are equal (w_j = 1), the formula recovers the unweighted circular mean. This construction preserves rotational invariance: adding a constant angle \alpha to all \theta_j shifts \mu by \alpha without altering \bar{R}. The concentration parameter \bar{R} provides a dispersion metric analogous to $1 - \bar{R} for variance in linear statistics, with higher values indicating greater clustering around \mu. In practice, the weighted circular mean addresses scenarios with unequal observation precisions, such as meta-analyses of directional where weights reflect sample sizes or errors from multiple sources. For instance, combining orientation estimates from biological studies (e.g., paths) weights larger or more precise samples more heavily to yield a reliable overall . Under the of independent observations from a with fixed weights, (MLE) yields \hat{\mu} as the weighted circular mean, while the concentration parameter \kappa is obtained by iteratively solving A(\kappa) = \bar{R}, where A(\kappa) = I_1(\kappa)/I_0(\kappa) involves modified Bessel functions of the first kind; weights are normalized by their sum during computation to ensure the estimator's consistency.

Multidimensional means

The circular mean generalizes to higher dimensions as the mean direction on the hypersphere S^{d-1}, where data points are unit vectors in \mathbb{R}^d. In three dimensions, the spherical mean corresponds to the direction of the resultant unit vector on S^2, computed as \boldsymbol{\mu} = \frac{\sum_{i=1}^n \mathbf{x}_i}{\|\sum_{i=1}^n \mathbf{x}_i\|}, where each \mathbf{x}_i is a unit vector representing a direction. This \boldsymbol{\mu} can be expressed in spherical coordinates (\theta, \phi), with \theta = \arccos(\mu_z) as the polar angle and \phi = \arctan(\mu_y / \mu_x) as the azimuthal angle, providing the central tendency of directional data such as wind vectors or molecular orientations. For the general d-dimensional case, the hyperspherical mean on S^{d-1} is similarly defined as the normalized sum of unit vectors, \boldsymbol{\mu} = \frac{\bar{\mathbf{R}}}{\|\bar{\mathbf{R}}\|}, where \bar{\mathbf{R}} = \frac{1}{n} \sum_{i=1}^n \mathbf{x}_i is the mean resultant . The angular between points on the hypersphere is measured using the , given by \delta(\boldsymbol{\mu}, \mathbf{x}_i) = \arccos(\boldsymbol{\mu}^\top \mathbf{x}_i), which quantifies around the mean. This formulation arises as the under the Riemannian metric on the hypersphere, minimizing the sum of squared distances. Key properties of the hyperspherical mean include equivariance under orthogonal transformations: if all data points are transformed by an Q \in O(d), the mean transforms as Q \boldsymbol{\mu}, preserving the geometric structure. Additionally, the concentration of the is assessed via the norm of the resultant \|\bar{\mathbf{R}}\|, which approaches 1 for highly concentrated distributions and 0 for dispersion, serving as a measure analogous to the circular variance in . Estimation of the mean is closely tied to parametric models in directional statistics, such as the von Mises-Fisher for directional (where the maximum likelihood estimator for the mean parameter is the sample hyperspherical mean) and the Bingham distribution for axial (where the principal axis of concentration is given by the eigenvector corresponding to the eigenvalue of largest magnitude in the concentration matrix). Recent post-2020 advances in high-dimensional settings leverage Riemannian optimization techniques, such as accelerated gradient methods on the hypersphere, to efficiently compute the for large-scale data where direct summation becomes computationally prohibitive due to dimensionality. These methods, including global acceleration schemes, improve convergence rates over classical iterative averaging, enabling applications in tasks like on high-dimensional spheres.

References

  1. [1]
    None
    ### Summary of Directional Statistics
  2. [2]
    circmean — SciPy v1.16.2 Manual
    ### Summary of scipy.stats.circmean
  3. [3]
    Circular data in biology: advice for effectively implementing ...
    Jul 11, 2018 · When researchers expect a unimodal departure and cannot a priori define the expected mean direction, the Rayleigh test should be used. In case ...Missing: meteorology | Show results with:meteorology
  4. [4]
    Modelling circular time series - ScienceDirect.com
    Many scientific fields have applications in which circular data are collected and statistically analysed. ... (mean direction) and υ is a non-negative ...
  5. [5]
    Directional Statistics - Wiley Online Library
    Presents new and up-dated material on both the underlying theory and the practical methodology of directional statistics.
  6. [6]
    One Direction? A Tutorial for Circular Data Analysis Using R With ...
    In the previous section we have seen that the computation of a circular mean differs from that of a linear mean. ... unit vector u composed of the sine and ...
  7. [7]
    [PDF] Directional Statistics - palaeo
    May 3, 2016 · ... Directional St at istics. Kanti V. Mardia. Peter E. Jupp. University of Leeds. UK. Uradttersity of St Andrews, UK. JOHN WILEY 8c SONS, LTD.
  8. [8]
    [PDF] To the vector belong the spoils: Circular statistics in Stata
    arithmetic mean is likely to be a poor summary: the mean of 1◦ and 359◦ cannot sensibly be. 180◦. The solution is to use the vector mean direction as circular ...
  9. [9]
    [PDF] The circular statistics toolbox for Matlab - MPG.PuRe
    The length of the mean resultant vector is a crucial quantity for the measurement of circular spread or hypothesis ... mean resultant vector length of the s ...
  10. [10]
    Statistical Analysis of Circular Data
    Statistical Analysis of Circular Data. Statistical Analysis of Circular Data ... PDF; Export citation. Select 6 - Correlation and regression. 6 - Correlation ...
  11. [11]
    [PDF] Application of circular statistics to life science - DergiPark
    Mar 30, 2019 · Therefore, the mean direction is undefined in the circular uniform distribution. ... data has a uniform distribution on the perimeter of circle.
  12. [12]
  13. [13]
    [PDF] Tutorial 2 - Circular descriptives
    A mean value in circular data is referred to as the mean direction. For example, the mean direction in a dataset with observations on a compass pointing north ( ...Missing: numerical | Show results with:numerical
  14. [14]
    Circular Regression Trees and Forests with an Application to ...
    The circular response variable that is considered in this application is a 10-min average of wind direction measurements at Innsbruck Airport and Vienna ...
  15. [15]
    A Circular-Linear Probabilistic Model Based on Nonparametric ...
    Abstract. The joint probability density function of wind speed and wind direction serves as the mathematical basis for directional wind energy assessment.
  16. [16]
    [PDF] The measurement of mean wind, variances and covariances ... - AMT
    Jun 13, 2022 · To obtain an accurate mean wind speed and mean wind direction on a moving car, it is necessary to correct for flow distortion and remove the ...Missing: navigation | Show results with:navigation
  17. [17]
    A new approach to evaluate multimodal orientation behaviour of ...
    Nov 1, 2013 · After initial treatment of individual test data it is often valuable to compare mean orientation between groups. The conventional procedures of ...
  18. [18]
    Head direction cells in a migratory bird prefer north - Science
    Feb 4, 2022 · Black vector originating from the center depicts the mean vector with a circular mean direction (CMD) and mean vector length (MVL) as a ...Missing: applications | Show results with:applications
  19. [19]
    Emlen funnel experiments revisited: methods update for studying ...
    ... average filter before being used for orientation evaluation. ... Each dot outside the unit circle indicates the mean orientation of a single bird for all three ...
  20. [20]
    A New Method of Investigation of the Orientation of Galaxies ... - MDPI
    Apr 12, 2023 · It is based on the statistical analysis of the distribution of angles, giving the spatial orientation of galaxies in space. In order to obtain ...
  21. [21]
    A circular statistical method for extracting rotation measures
    Abstract. We propose a new method for the extraction of rotation measures from spectral polarization data. The method is based on maximum likelihood analys.
  22. [22]
    application of an optimal statistic for characterizing relative orientations
    The Rayleigh statistic can be used to test for the non-uniformity of a set of angles, and has already been applied many times in astrophysics when testing for ...Missing: positional | Show results with:positional
  23. [23]
    Robust beamforming in circular arrays using phase‐mode ...
    Oct 1, 2013 · Uniform circular array (UCA) is a structure frequently used in practical applications such as sonar, radar and wireless communications [1].
  24. [24]
    [PDF] Using Signal Phase in Computerized Systems of Non-destructive ...
    Circular statistics, such as the mean ring value, circular variance and the resulting vector length, were proposed for use in the phase measurements. Some ...<|separator|>
  25. [25]
    [PDF] circular dataimages for directional data
    Historical Summary (Fisher, 1993). Most of the theoretical advances in the statistics of circularhxial random variables are relatively recent. In 1767, John ...
  26. [26]
    [PDF] Fisher's Legacy of Directional Statistics, and Beyond to ... - arXiv
    May 28, 2024 · We will describe this Fisher distribution and reanalyse his geological data. We also comment on the two goals he set himself in that paper, and ...
  27. [27]
  28. [28]
    An algorithm for computing Fréchet means on the sphere
    Mar 7, 2019 · We present an algorithm for solving a special global problem over the sphere, namely the determination of Fréchet means, which are points minimising the mean ...