LibBi is used for state-space modelling and Bayesian inference on high-performance computer hardware, including multi-core CPUs, many-core GPUs (graphics processing units), and distributed-memory clusters. The staple methods of LibBi are based on sequential Monte Carlo (SMC), also known as particle filtering. These methods include particle Markov chain Monte Carlo (PMCMC) and SMC2. Other methods include the extended Kalman filter and some parameter optimization routines. LibBi consists of a C++ template library and a parser and compiler, written in Perl, for its own modelling language.
MLPACK is a C++ machine learning library with an emphasis on scalability, speed, and ease-of-use. Its aim is to make machine learning possible for novice users by means of a simple, consistent API, while simultaneously exploiting C++ language features to provide maximum performance and maximum flexibility for expert users. It contains algorithms such as k-means, Gaussian mixture models, hidden Markov models, density estimation trees, kernel PCA, locality-sensitive hashing, sparse coding, linear regression and least-angle regression.
Meta.Numerics is a Mono-compatible .NET library for scientific and numerical programming. It includes functionality for matrix algebra (including SVD, non-symmetric eigensystems, and sparse matrices), special functions of real and complex numbers (including Bessel functions and the complex error function), statistics and data analysis (including PCA, logistic and nonlinear regression, statistical tests, and nonuniform random deviates), and signal processing (including arbitrary-length FFTs).
Metrix++ is a platform to collect and analyze code metrics. It has a plugin-based architecture, so it is easy to add support for new languages, define new metrics, and/or create new pre- and post-processing tools. Every metric has 'turn-on' and other configuration options. There are no predefined thresholds for metrics or rules; you can choose and configure any limit you want. It scales well to large codebases. For example, initial parsing of about 10000 files takes 2-3 minutes on an average PC, and only 10-20 seconds for iterative re-run. Reporting summary results and exceeded limits takes less than 1 - 10 seconds. It can compare results for 2 code snapshots (collections) and differentiate added regions (classes, functions, etc.), modified regions, and unchanged regions. As a result, easy deployment is guaranteed into legacy software, helping you to deal with legacy code efficiently, and either enforce the 'leave it not worse than it was before' rule or motivate re-factoring.
PHP Clarke and Wright Algorithm is a class that can solve a truck routing problem with the Clarke and Wright algorithm. It attempts to solve the problem of determining the routes by which a given number of trucks with different weight and volume capacity will be dispatching deliveries to a certain number of clients distributed geographically within certain time windows. The class takes as parameters the nodes of positions of each client, the demands of each client, a matrix of distance between nodes, and the capacity of each truck. It computes the route for each truck, as well the time and distance to drive to each customer and the volume and weight to transport.