Flexible and consistent quantile estimation for intensity–duration–frequency curves

Assessing the relationship between the intensity, duration, and frequency (IDF) of extreme precipitation is required for the design of water management systems. However, when modeling sub-daily precipitation extremes, there are commonly only short observation time series available. This problem can be overcome by applying the duration-dependent formulation of the generalized extreme value (GEV) distribution which fits an IDF model with a range of durations simultaneously. The originally proposed duration-dependent GEV model exhibits a power-law-like behavior of the quantiles and takes care of a deviation from this scaling relation (curvature) for sub-hourly durations (Koutsoyiannis et al., 1998). We suggest that a more flexible model might be required to model a wide range of durations (1 min to 5 d). Therefore, we extend the model with the following two features: (i) different slopes for different quantiles (multiscaling) and (ii) the deviation from the power law for large durations (flattening), which is newly introduced in this study. Based on the quantile skill score, we investigate the performance of the resulting flexible model with respect to the benefit of the individual features (curvature, multiscaling, and flattening) with simulated and empirical data. We provide detailed information on the duration and probability ranges for which specific features or a systematic combination of features leads to improvements for stations in a case study area in the Wupper catchment (Germany). Our results show that allowing curvature or multiscaling improves the model only for very short or long durations, respectively, but leads to disadvantages in modeling the other duration ranges. In contrast, allowing flattening on average leads to an improvement for medium durations between 1 h and 1 d, without affecting other duration regimes. Overall, the new parametric form offers a flexible and enhanced performance model for consistently describing IDF relations over a wide range of durations, which has not been done before as most existing studies focus on durations longer than 1 h or day and do not address the deviation from the power law for very long durations (2–5 d).

The stream of life: from frozen babies to the birth of psychic life

This article describes the work with a flexible model having Esther Bick’s infant observation model in mind. Having to solve the problem of the many babies in orphanages who had been abandoned and were often sick, I worked out with the psychologist a model of intervention using the knowledge the psychologist had of the approach of infant observation, but adapting it to the context of the orphanage. I helped the psychologist observer to use her countertransference to contain the babies’ distress and interact with them in order to give them the possibility of feeling loved and understood, in order to make space in their minds for introjecting a good object. On behalf of the observations that the psychologist, whom I will call Ling, had with babies Bea and Sally, we can say that the dramatic distress of their being abandoned by their families has been worked through because the psychologist was able to give a loving space with significant interaction. The sessions reported in this article give a portrait of the inside world of despairing babies, and document the evolution towards being able to keep in mind a good enough object that can help them survive.

The McDonald Lindley-Poisson Distribution

We propose the McDonald Lindley-Poisson distribution and derive some of its mathematical properties including explicit expressions for moments, generating and quantile functions, mean deviations, order statistics and their moments. Its model parameters are estimated by maximum likelihood. A simulation study investigates the performance of the estimates. The new distribution represents a more flexible model for lifetime data analysis than other existing models as proved empirically by means of two real data sets.

Deep embedding and alignment of protein sequences

Protein sequence alignment is a key component of most bioinformatics pipelines to study the structures and functions of proteins. Aligning highly divergent sequences remains, however, a difficult task that current algorithms often fail to perform accurately, leaving many proteins or open reading frames poorly annotated. Here, we leverage recent advances in deep learning for language modelling and differentiable programming to propose DEDAL, a flexible model to align protein sequences and detect homologs. DEDAL is a machine learning-based model that learns to align sequences by observing large datasets of raw protein sequences and of correct alignments. Once trained, we show that DEDAL improves by up to two- or three-fold the alignment correctness over existing methods on remote homologs, and better discriminates remote homologs from evolutionarily unrelated sequences, paving the way to improvements on many downstream tasks relying on sequence alignment in structural and functional genomics.

Does proximity to conflict affect tourism: Evidence from NATO bombing

Although conflict, war, violence, and terrorism affect tourism, research that identifies possible channels of these effects is scarce. We explore if the adverse effects are channelled through proximity to conflict areas. We use the conflict in Kosovo in 1999 and the country Croatia as a quasi-natural experiment and take advantage of the specific north-west to south-east orientation of Croatian Adriatic counties to identify the effect of NATO bombing in Kosovo on tourism outcomes as well as the potential proximity channel. Using data on the population of Croatian firms and the difference-in-differences identification strategy we find that tourism companies’ revenues decreased significantly due to NATO bombing, especially in accommodation services and in companies with 50 or more employees. However, using a synthetic control approach we find that the adverse effect is only transitory. Analysing heterogeneous effects with respect to the distance of the firm from Kosovo—using a linear and a more flexible model—we find compelling evidence that within-country proximity to conflict is not a significant channel through which the negative effect propagates.

Bayesian tests of symmetry for the generalized Von Mises distribution

Bayesian tests on the symmetry of the generalized von Mises model for planar directions (Gatto and Jammalamadaka in Stat Methodol 4(3):341–353, 2007) are introduced. The generalized von Mises distribution is a flexible model that can be axially symmetric or asymmetric, unimodal or bimodal. A characterization of axial symmetry is provided and taken as null hypothesis for one of the proposed Bayesian tests. The Bayesian tests are obtained by the technique of probability perturbation. The prior probability measure is perturbed so to give a positive prior probability to the null hypothesis, which would be null otherwise. This allows for the derivation of simple computational formulae for the Bayes factors. Numerical results reveal that, whenever the simulation scheme of the samples supports the null hypothesis, the null posterior probabilities appear systematically larger than their prior counterpart.

Modeling Proportion Data with Inflation by Using a Power-Skew-Normal/Logit Mixture Model

Rate or proportion data are modeled by using a regression model. The considered regression model can be used for studying phenomena with a response on the (0, 1), [0, 1), (0, 1], or [0, 1] intervals. To connect the response variable with the linear predictor in the regression model, we use a logit link function, which guarantees that the obtained prediction ranges between zero and one in the cases inflated at zero or one (or both). The model is complemented with the assumption that the errors follow a power-skew-normal distribution, resulting in a very flexible model, and with a non-singular information matrix, constituting an advantage over other existing models in the literature. To explain the probability of point mass at the values zero and/or one (inflated part), we used a polytomic logistic model with covariates. The results of two illustrations showed that the proposed model is a better alternative compared to widely known models in the literature.