Umeå universitets logga

This paper presents an R package, fdaMocca, that provides routines for model-based functional cluster analysis for functional data with optional covariates. The idea is to cluster functional subjects (often called functional objects) into homogenous groups by using spline smoothers (for functional data) together with scalar covariates. The spline coefficients and the covariates are modelled as a multivariate Gaussian mixture model, where the number of mixtures corresponds to the number of clusters. The parameters of the model are estimated by maximizing the observed mixture likelihood via an EM algorithm. The usefulness of fdaMocca and its clustering methods is illustrated on a functional data set with covariates from 6400 annual seasonal patterns of varved lake sediment from Lake Kassjön (Northern Sweden). Each varve contains information about the weather the year the varve was formed and thus may be used to reconstruct past climate.

Functional regression is a dynamic research field within functional data analysis, with a wide range of applications across different areas. This paper aims to expand the existing framework of shape-constrained generalized additive models to functional generalized additive models with shape constraints. We introduce an extension of the shape-constrained P-spline (SCOP-spline) approach to a broad class of functional regression models with various shape constraints. Our framework includes parametric and a mixture of constrained and unconstrained smooth effects of functional and scalar covariates. Estimation and inference in this framework build upon the shape-constrained generalized additive models, enabling the use of wellestablished, robust, and flexible procedures. The described methods are implemented in the user-friendly R package scam. Simulation shows performance improvements when modelling with the proposed approach compared to the unconstrained method.

We have developed a homemade neutron flux detection module with 3He tube hot-swap capability and control-rich Android software interface. Real-time data analysis is done by a smartphone with Android application interfaced with the detector via a USB cable. This setup can be used as a neutron and gamma ray background detector or as a compact, mobile 3He tubes calibration tool making it a cheap and easy-to-use alternative for the stationary setups. A fast neutron detection algorithm was implemented as a set of Java scripts and tested for real-time signal analysis. The modular structure of the device allows easy deployment and customization with further software development and regular upgrades. The current prototype was tested at the Nuclear Physics Research Institute under different neutron flux intensity conditions from the VVR-K water-cooled research reactor. Its simplicity and significantly lower cost, compared with conventional detector equipment, make it valuable for easy repetitive tasks with medium requirements for precision and neutron flux intensities.

Regression models that incorporate smooth functions of predictor variables to explain the relationships with a response variable have gained widespread usage and proved successful in various applications. By incorporating smooth functions of predictor variables, these models can capture complex relationships between the response and predictors while still allowing for interpretation of the results. In situations where the relationships between a response variable and predictors are explored, it is not uncommon to assume that these relationships adhere to certain shape constraints. Examples of such constraints include monotonicity and convexity. The scam package for R has become a popular package to carry out the full fitting of exponential family generalized additive modelling with shape restrictions on smooths. The paper aims to extend the existing framework of shape-constrained generalized additive models (SCAM) to accommodate smooth interactions of covariates, linear functionals of shape-constrained smooths and incorporation of residual autocorrelation. The methods described in this paper are implemented in the recent version of the package scam, available on the Comprehensive R Archive Network (CRAN).

scam provides functions for generalized additive modelling under shape constraints on the component functions of the linear predictor of the GAM. Models can contain multiple shape constrained and unconstrained terms as well as bivariate smooths with double or single monotonicity.

We introduce an algorithm which can be directly used to feasible and optimum search in linear programming. Starting from an initial point the algorithm iteratively moves a point in a direction to resolve the violated constraints. At the same time, it ensures that previously fulfilled constraints are not breached during this process. The method is based on geometrical properties of n-dimensional space and can be used on any type of linear constraints (>, =, ≥), moreover it can be used when the feasible region is non-full-dimensional.