Search

Classical tests are available for the two-sample test of correspondence of distribution functions. From these, the Kolmogorov-Smirnov test provides also the graphical interpretation of the test results, in different forms. Here, we propose modifications of the Kolmogorov-Smirnov test with higher power. The proposed tests are based on the so-called global envelope test which allows for graphical interpretation, similarly as the Kolmogorov-Smirnov test. The tests are based on rank statistics and are suitable also for the comparison of $n$ samples, with $n \geq 2$. We compare the alternatives for the two-sample case through an extensive simulation study and discuss their interpretation. Finally, we apply the tests to real data. Specifically, we compare the height distributions between boys and girls at different ages, the sepal length distributions of different flower species, and distributions of standardized residuals from a time series model for different exchange courses using the proposed methodologies., Comment: 23 pages, 17 figures

Many natural systems are observed as point patterns in time, space, or space and time. Examples include plant and cellular systems, animal colonies, earthquakes, and wildfires. In practice the locations of the points are not always observed correctly. However, in the point process literature, there has been relatively scant attention paid to the issue of errors in the location of points. In this paper, we discuss how the observed point pattern may deviate from the actual point pattern and review methods and models that exist to handle such deviations. The discussion is supplemented with several scientific illustrations.

Quantile regression is used to study effects of covariates on a particular quantile of the data distribution. Here we are interested in the question whether a covariate has any effect on the entire data distribution, i.e., on any of the quantiles. To this end, we treat all the quantiles simultaneously and consider global tests for the existence of the covariate effect in the presence of nuisance covariates. This global quantile regression can be used as the extension of linear regression or as the extension of distribution comparison in the sense of Kolmogorov-Smirnov test. The proposed method is based on pointwise coefficients, permutations and global envelope tests. The global envelope test serves as the multiple test adjustment procedure under the control of the family-wise error rate and provides the graphical interpretation which automatically shows the quantiles or the levels of categorical covariate responsible for the rejection. The Freedman-Lane permutation strategy showed liberality of the test for extreme quantiles, therefore we propose four alternatives that work well even for extreme quantiles and are suitable in different conditions. We present a simulation study to inspect the performance of these strategies, and we apply the chosen strategies to two data examples., Comment: 44 pages, 12 figures

Determinantal point processes are models for regular spatial point patterns, with appealing probabilistic properties. We present their spatio-temporal counterparts and give examples of these models, based on spatio-temporal covariance functions which are separable and non-separable in space and time., Comment: 9 pages, 1 figure

Light detection and ranging (LiDAR) provides information on the vertical structure of forest stands enabling detailed and extensive ecosystem study. The vertical structure is often summarized by scalar features and data-reduction techniques that limit the interpretation of results. Instead, we quantified the influence of three variables, species, crown cover, and age, on the vertical distribution of airborne LiDAR returns from forest stands. We studied 5,428 regular, even-aged stands in Quebec (Canada) with five dominant species: balsam fir (Abies balsamea (L.) Mill.), paper birch (Betula papyrifera Marsh), black spruce (Picea mariana (Mill.) BSP), white spruce (Picea glauca Moench) and aspen (Populus tremuloides Michx.). We modeled the vertical distribution against the three variables using a functional general linear model and a novel nonparametric graphical test of significance. Results indicate that LiDAR returns from aspen stands had the most uniform vertical distribution. Balsam fir and white birch distributions were similar and centered at around 50% of the stand height, and black spruce and white spruce distributions were skewed to below 30% of stand height (p<0.001). Increased crown cover concentrated the distributions around 50% of stand height. Increasing age gradually shifted the distributions higher in the stand for stands younger than 70-years, before plateauing and slowly declining at 90-120 years. Results suggest that the vertical distributions of LiDAR returns depend on the three variables studied.

First-order separability of a spatio-temporal point process plays a fundamental role in the analysis of spatio-temporal point pattern data. While it is often a convenient assumption that simplifies the analysis greatly, existing non-separable structures should be accounted for in the model construction. We propose three different tests to investigate this hypothesis as a step of preliminary data analysis. The first two tests are exact or asymptotically exact for Poisson processes. The first test based on permutations and global envelopes allows us to detect at which spatial and temporal locations or lags the data deviate from the null hypothesis. The second test is a simple and computationally cheap $\chi^2$-test. The third test is based on statistical reconstruction method and can be generally applied for non-Poisson processes. The performance of the first two tests is studied in a simulation study for Poisson and non-Poisson models. The third test is applied to the real data of the UK 2001 epidemic foot and mouth disease., Comment: 21 pages, 8 Figures (21 plots)

False discovery rate (FDR) is a common way to control the number of false discoveries in multiple testing. There are a number of approaches available for controlling FDR. However, for functional test statistics, which are discretized into $m$ highly correlated hypotheses, the methods must account for changes in distribution across the functional domain and correlation structure. Further, it is of great practical importance to visualize the test statistic together with its rejection or acceptance region. Therefore, the aim of this paper is to find, based on resampling principles, a graphical envelope that controls FDR and detects the outcomes of all individual hypotheses by a simple rule: the hypothesis is rejected if and only if the empirical test statistic is outside of the envelope. Such an envelope offers a straightforward interpretation of the test results, similarly as the recently developed global envelope testing which controls the family-wise error rate. Two different adaptive single threshold procedures are developed to fulfill this aim. Their performance is studied in an extensive simulation study. The new methods are illustrated by three real data examples.

This study presents a simulation study to compare different non-parametric global envelopes that are refinements of the rank envelope proposed by Myllym\"aki et al. (2017, Global envelope tests for spatial processes, J. R. Statist. Soc. B 79, 381-404, doi: 10.1111/rssb.12172). The global envelopes are constructed for a set of functions or vectors. For a large number of vectors, all the refinements lead to the same outcome as the global rank envelope. For smaller numbers of vectors the refinement playes a role, where different refinements are sensitive to different types of extremeness of a vector among the set of vectors. The performance of the different alternatives are compared in a simulation study with respect to the numbers of available vectors, the dimensionality of the vectors, the amount of dependence between the vector elements and the expected type of extremeness.

We propose a hierarchical log Gaussian Cox process (LGCP) for point patterns, where a set of points x affects another set of points y but not vice versa. We use the model to investigate the effect of large trees to the locations of seedlings. In the model, every point in x has a parametric influence kernel or signal, which together form an influence field. Conditionally on the parameters, the influence field acts as a spatial covariate in the intensity of the model, and the intensity itself is a non-linear function of the parameters. Points outside the observation window may affect the influence field inside the window. We propose an edge correction to account for this missing data. The parameters of the model are estimated in a Bayesian framework using Markov chain Monte Carlo (MCMC) where a Laplace approximation is used for the Gaussian field of the LGCP model. The proposed model is used to analyze the effect of large trees on the success of regeneration in uneven-aged forest stands in Finland., Comment: 28 pages, 9 figures

The aim of the paper is to construct spatial models for the activation of sweat glands for healthy subjects and subjects suffering from peripheral neuropathy by using videos of sweating recorded from the subjects. The sweat patterns are regarded as realizations of spatial point processes and two point process models for the sweat gland activation and two methods for inference are proposed. Several image analysis steps are needed to extract the point patterns from the videos and some incorrectly identified sweat gland locations may be present in the data. To take into account the errors we either include an error term in the point process model or use an estimation procedure that is robust with respect to the errors., Comment: 27 pages, 12 figures

This work describes the R package GET that implements global envelopes for a general set of $d$-dimensional vectors $T$ in various applications. A $100(1-\alpha)$% global envelope is a band bounded by two vectors such that the probability that $T$ falls outside this envelope in any of the $d$ points is equal to $\alpha$. The term 'global' means that this probability is controlled simultaneously for all the $d$ elements of the vectors. The global envelopes can be employed for central regions of functional or multivariate data, for graphical Monte Carlo and permutation tests where the test statistic is multivariate or functional, and for global confidence and prediction bands. Intrinsic graphical interpretation property is introduced for global envelopes. The global envelopes included in the GET package have this property, which particularly helps to interpret test results, by providing a graphical interpretation that shows the reasons of rejection of the tested hypothesis. Examples of different uses of global envelopes and their implementation in the GET package are presented, including global envelopes for single and several one- or two-dimensional functions, Monte Carlo goodness-of-fit tests for simple and composite hypotheses, comparison of distributions, functional analysis of variance, functional linear model, and confidence bands in polynomial regression.

With recent developments in remote sensing technologies, plot-level forest resources can be predicted utilizing airborne laser scanning (ALS). The prediction is often assisted by mostly vertical summaries of the ALS point clouds. We present a spatial analysis of the point cloud by studying the horizontal distribution of the pulse returns through canopy height models thresholded at different height levels. The resulting patterns of patches of vegetation and gabs on each layer are summarized to spatial ALS features. We propose new features based on the Euler number, which is the number of patches minus the number of gaps, and the empty-space function, which is a spatial summary function of the gab space. The empty-space function is also used to describe differences in the gab structure between two different layers. We illustrate usefulness of the proposed spatial features for predicting different forest variables that summarize the spatial structure of forests or their breast height diameter distribution. We employ the proposed spatial features, in addition to commonly used features from literature, in the well-known k-nn estimation method to predict the forest variables. We present the methodology on the example of a study site in Central Finland., Comment: 22 pages, 5 figures, 5 tables (including the appendix). Taking valuable comments on our manuscript [v2] into consideration, we have decided to emphasize the focus on the spatial analysis of airborne laser scanning (ALS) point clouds and the new spatial ALS feature variables. By doing so, we have reduced the methodology for the spatial structure of forests evaluated on the field plot level

The rank envelope test (Myllym\"aki et al., Global envelope tests for spatial processes, arXiv:1307.0239 [stat.ME]) is proposed as a solution to multiple testing problem for Monte Carlo tests. Three different situations are recognized: 1) a few univariate Monte Carlo tests, 2) a Monte Carlo test with a function as the test statistic, 3) several Monte Carlo tests with functions as test statistics. The rank test has correct (global) type I error in each case and it is accompanied with a $p$-value and with a graphical interpretation which shows which subtest or which distances of the used test function(s) lead to the rejection at the prescribed significance level of the test. Examples of null hypothesis from point process and random set statistics are used to demonstrate the strength of the rank envelope test. The examples include goodness-of-fit test with several test functions, goodness-of-fit test for one group of point patterns, comparison of several groups of point patterns, test of dependence of components in a multi-type point pattern, and test of Boolean assumption for random closed sets.

Envelope tests are a popular tool in spatial statistics, where they are used in goodness-of-fit testing. These tests graphically compare an empirical function $T(r)$ with its simulated counterparts from the null model. However, the type I error probability $\alpha$ is conventionally controlled for a fixed distance $r$ only, whereas the functions are inspected on an interval of distances $I$. In this study, we propose two approaches related to Barnard's Monte Carlo test for building global envelope tests on $I$:(1) ordering the empirical and simulated functions based on their $r$-wise ranks among each other, and (2) the construction of envelopes for a deviation test. These new tests allow the a priori selection of the global $\alpha$ and they yield $p$-values. We illustrate these tests using simulated and real point pattern data.

The deviation test belong to core tools in point process statistics, where hypotheses are typically tested considering differences between an empirical summary function and its expectation under the null hypothesis, which depend on a distance variable r. This test is a classical device to overcome the multiple comparison problem which appears since the functional differences have to be considered for a range of distances r simultaneously. The test has three basic ingredients: (i) choice of a suitable summary function, (ii) transformation of the summary function or scaling of the differences, and (iii) calculation of a global deviation measure. We consider in detail the construction of such tests both for stationary and finite point processes and show by two toy examples and a simulation study for the case of the random labelling hypothesis that the points (i) and (ii) have great influence on the power of the tests.