Data Structures, Summaries, and Visualisations for Missing Data

Missing values are ubiquitous in data and need to be explored and handled in the initial stages of analysis. 'naniar' provides data structures and functions that facilitate the plotting of missing values and examination of imputations. This allows missing data dependencies to be explored with minimal deviation from the common work patterns of 'ggplot2' and tidy data. The work is fully discussed at Tierney & Cook (2023) .

Get XKCD Comic Data

Download data from individual XKCD comics, written by Randall Munroe < https://xkcd.com/>.

Spatial Temporal Analysis of Moving Polygons

Perform spatial temporal analysis of moving polygons; a longstanding analysis problem in Geographic Information Systems. Facilitates directional analysis, distance analysis, and some other simple functionality for examining spatial-temporal patterns of moving polygons.

Automatic Fixed Rank Kriging

Automatic fixed rank kriging for (irregularly located) spatial data using a class of basis functions with multi-resolution features and ordered in terms of their resolutions. The model parameters are estimated by maximum likelihood (ML) and the number of basis functions is determined by Akaike's information criterion (AIC). For spatial data with either one realization or independent replicates, the ML estimates and AIC are efficiently computed using their closed-form expressions when no missing value occurs. Details regarding the basis function construction, parameter estimation, and AIC calculation can be found in Tzeng and Huang (2018) . For data with missing values, the ML estimates are obtained using the expectation- maximization algorithm. Apart from the number of basis functions, there are no other tuning parameters, making the method fully automatic. Users can also include a stationary structure in the spatial covariance, which utilizes 'LatticeKrig' package.

Weighted Portmanteau Tests for Time Series Goodness-of-Fit

An implementation of the Weighted Portmanteau Tests described in "New Weighted Portmanteau Statistics for Time Series Goodness-of-Fit Testing" published by the Journal of the American Statistical Association, Volume 107, Issue 498, pages 777-787, 2012.

Overdispersion in Count Data Multiple Regression Analysis

Detection of overdispersion in count data for multiple regression analysis. Log-linear count data regression is one of the most popular techniques for predictive modeling where there is a non-negative discrete quantitative dependent variable. In order to ensure the inferences from the use of count data models are appropriate, researchers may choose between the estimation of a Poisson model and a negative binomial model, and the correct decision for prediction from a count data estimation is directly linked to the existence of overdispersion of the dependent variable, conditional to the explanatory variables. Based on the studies of Cameron and Trivedi (1990) and Cameron and Trivedi (2013, ISBN:978-1107667273), the overdisp() command is a contribution to researchers, providing a fast and secure solution for the detection of overdispersion in count data. Another advantage is that the installation of other packages is unnecessary, since the command runs in the basic R language.

Censoring Data and Likelihood-Based Correlation Estimation

A profile likelihood based method of estimation and inference on the correlation coefficient of bivariate data with different types of censoring and missingness.

Spectral Modularity Clustering

Implements the network clustering algorithm described in Newman (2006) . The complete iterative algorithm comprises of two steps. In the first step, the network is expressed in terms of its leading eigenvalue and eigenvector and recursively partition into two communities. Partitioning occurs if the maximum positive eigenvalue is greater than the tolerance (10e-5) for the current partition, and if it results in a positive contribution to the Modularity. Given an initial separation using the leading eigen step, 'rSpectral' then continues to maximise for the change in Modularity using a fine-tuning step - or variate thereof. The first stage here is to find the node which, when moved from one community to another, gives the maximum change in Modularity. This node’s community is then fixed and we repeat the process until all nodes have been moved. The whole process is repeated from this new state until the change in the Modularity, between the new and old state, is less than the predefined tolerance. A slight variant of the fine-tuning step, which can improve speed of the calculation, is also provided. Instead of moving each node into each community in turn, we only consider moves of neighbouring nodes, found in different communities, to the community of the current node of interest. The two steps process is repeatedly applied to each new community found, subdivided each community into two new communities, until we are unable to find any division that results in a positive change in Modularity.

Companion Software for the Coursera Statistics with R Specialization

Data and functions to support Bayesian and frequentist inference and decision making for the Coursera Specialization "Statistics with R". See < https://github.com/StatsWithR/statsr> for more information.

This is a Collection of Functions to Analyse Gender Differences

Implementation of functions, which combines binomial calculation and data visualisation, to analyse the differences in publishing authorship by gender described in Day et al. (2020) . It should only be used when self-reported gender is unavailable.