Do algebraic operations on neural networks. We seek here to implement in R, operations on neural networks and their resulting approximations. Our operations derive their descriptions mainly from Rafi S., Padgett, J.L., and Nakarmi, U. (2024), "Towards an Algebraic Framework For Approximating Functions Using Neural Network Polynomials", <doi:10.48550/arXiv.2402.01058>, Grohs P., Hornung, F., Jentzen, A. et al. (2023), "Space-time error estimates for deep neural network approximations for differential equations", <doi:10.1007/s10444-022-09970-2>, Jentzen A., Kuckuck B., von Wurstemberger, P. (2023), "Mathematical Introduction to Deep Learning Methods, Implementations, and Theory" <doi:10.48550/arXiv.2310.20360>. Our implementation is meant mainly as a pedagogical tool, and proof of concept. Faster implementations with deeper vectorizations may be made in future versions.

We solve non linear least squares problems with optional equality and/or inequality constraints. Non linear iterations are globalized with back-tracking method. Linear problems are solved by dense QR decomposition from LAPACK which can limit the size of treated problems. On the other side, we avoid condition number degradation which happens in classical quadratic programming approach. Inequality constraints treatment on each non linear iteration is based on NNLS method (by Lawson and Hanson). We provide an original function lsi_ln for solving linear least squares problem with inequality constraints in least norm sens. Thus if Jacobian of the problem is rank deficient a solution still can be provided. However, truncation errors are probable in this case. Equality constraints are treated by using a basis of Null-space. User defined function calculating residuals must return a list having residual vector (not their squared sum) and Jacobian. If Jacobian is not in the returned list, package numDeriv is used to calculated finite difference version of Jacobian. The NLSIC method was fist published in Sokol et al. (2012) <doi:10.1093/bioinformatics/btr716>.

This package implements calculation of probability density function, cumulative distribution function, equicoordinate quantile function and survival function, and random numbers generation for the following multivariate distributions: Lomax (Pareto Type II), generalized Lomax, Mardiaâ s Pareto of Type I, Logistic, Burr, Cook-Johnsonâ s uniform, F and Inverted Beta. See Tapan Nayak (1987) <doi:10.2307/3214068>.

Fast functions implemented in C++ via Rcpp to support the NeuroAnatomy Toolbox ('nat') ecosystem. These functions provide large speed-ups for basic manipulation of neuronal skeletons over pure R functions found in the nat package. The expectation is that end users will not use this package directly, but instead the nat package will automatically use routines from this package when it is available to enable large performance gains.

This package implements likelihood inference based on higher order approximations for nonlinear models with possibly non constant variance.

This package provides tools for working with nonlinear least squares problems. For the estimation of models reliable and robust tools than nls(), where the the Gauss-Newton method frequently stops with singular gradient messages. This is accomplished by using, where possible, analytic derivatives to compute the matrix of derivatives and a stabilization of the solution of the estimation equations. Tools for approximate or externally supplied derivative matrices are included. Bounds and masks on parameters are handled properly.

We developed an inference tool based on approximate Bayesian computation to decipher network data and assess the strength of the inferred links between network's actors. It is a new multi-level approximate Bayesian computation (ABC) approach. At the first level, the method captures the global properties of the network, such as a scale-free structure and clustering coefficients, whereas the second level is targeted to capture local properties, including the probability of each couple of genes being linked. Up to now, Approximate Bayesian Computation (ABC) algorithms have been scarcely used in that setting and, due to the computational overhead, their application was limited to a small number of genes. On the contrary, our algorithm was made to cope with that issue and has low computational cost. It can be used, for instance, for elucidating gene regulatory network, which is an important step towards understanding the normal cell physiology and complex pathological phenotype. Reverse-engineering consists in using gene expressions over time or over different experimental conditions to discover the structure of the gene network in a targeted cellular process. The fact that gene expression data are usually noisy, highly correlated, and have high dimensionality explains the need for specific statistical methods to reverse engineer the underlying network.

Next-Generation Clustered Heat Maps (NG-CHMs) allow for dynamic exploration of heat map data in a web browser. NGCHM allows users to create both stand-alone HTML files containing a Next-Generation Clustered Heat Map, and .ngchm files to view in the NG-CHM viewer. See Ryan MC, Stucky M, et al (2020) <doi:10.12688/f1000research.20590.2> for more details.

Minimize a differentiable function subject to all the variables being non-negative (i.e. >= 0), using a Conjugate-Gradient algorithm based on a modified Polak-Ribiere-Polyak formula as described in (Li, Can, 2013, <https://www.hindawi.com/journals/jam/2013/986317/abs/>).

Includes assorted tools for network analysis. Bridge centrality; goldbricker; MDS, PCA, & eigenmodel network plotting.

This package contains the functions for testing the spatial patterns (of segregation, spatial symmetry, association, disease clustering, species correspondence, and reflexivity) based on nearest neighbor relations, especially using contingency tables such as nearest neighbor contingency tables (Ceyhan (2010) <doi:10.1007/s10651-008-0104-x> and Ceyhan (2017) <doi:10.1016/j.jkss.2016.10.002> and references therein), nearest neighbor symmetry contingency tables (Ceyhan (2014) <doi:10.1155/2014/698296>), species correspondence contingency tables and reflexivity contingency tables (Ceyhan (2018) <doi:10.2436/20.8080.02.72> for two (or higher) dimensional data. The package also contains functions for generating patterns of segregation, association, uniformity in a multi-class setting (Ceyhan (2014) <doi:10.1007/s00477-013-0824-9>), and various non-random labeling patterns for disease clustering in two dimensional cases (Ceyhan (2014) <doi:10.1002/sim.6053>), and for visualization of all these patterns for the two dimensional data. The tests are usually (asymptotic) normal z-tests or chi-square tests.

NEON data packages can be accessed through the NEON Data Portal <https://www.neonscience.org> or through the NEON Data API (see <https://data.neonscience.org/data-api> for documentation). Data delivered from the Data Portal are provided as monthly zip files packaged within a parent zip file, while individual files can be accessed from the API. This package provides tools that aid in discovering, downloading, and reformatting data prior to use in analyses. This includes downloading data via the API, merging data tables by type, and converting formats. For more information, see the readme file at <https://github.com/NEONScience/NEON-utilities>.

Fits regularization paths for linear regression, GLM, and Cox regression models using lasso or nonconvex penalties, in particular the minimax concave penalty (MCP) and smoothly clipped absolute deviation (SCAD) penalty, with options for additional L2 penalties (the "elastic net" idea). Utilities for carrying out cross-validation as well as post-fitting visualization, summarization, inference, and prediction are also provided. For more information, see Breheny and Huang (2011) <doi:10.1214/10-AOAS388> or visit the ncvreg homepage <https://pbreheny.github.io/ncvreg/>.

This package implements a method that builds the coefficients of a polynomial model that performs almost equivalently as a given neural network (densely connected). This is achieved using Taylor expansion at the activation functions. The obtained polynomial coefficients can be used to explain features (and their interactions) importance in the neural network, therefore working as a tool for interpretability or eXplainable Artificial Intelligence (XAI). See Morala et al. 2021 <doi:10.1016/j.neunet.2021.04.036>, and 2023 <doi:10.1109/TNNLS.2023.3330328>.

Various visual and numerical diagnosis methods for the nonlinear mixed effect model, including visual predictive checks, numerical predictive checks, and coverage plots (Karlsson and Holford, 2008, <https://www.page-meeting.org/?abstract=1434>).

This package implements the framework presented in Cucci, D. A., Voirol, L., Khaghani, M. and Guerrier, S. (2023) <doi:10.1109/TIM.2023.3267360> which allows to analyze the impact of sensor error modeling on the performance of integrated navigation (sensor fusion) based on inertial measurement unit (IMU), Global Positioning System (GPS), and barometer data. The framework relies on Monte Carlo simulations in which a Vanilla Extended Kalman filter is coupled with realistic and user-configurable noise generation mechanisms to recover a reference trajectory from noisy measurements. The evaluation of several statistical metrics of the solution, aggregated over hundreds of simulated realizations, provides reasonable estimates of the expected performances of the system in real-world conditions.

NanoString nCounter data are gene expression assays where there is no need for the use of enzymes or amplification protocols and work with fluorescent barcodes (Geiss et al. (2018) <doi:10.1038/nbt1385>). Each barcode is assigned a messenger-RNA/micro-RNA (mRNA/miRNA) which after bonding with its target can be counted. As a result each count of a specific barcode represents the presence of its target mRNA/miRNA. NACHO (NAnoString quality Control dasHbOard) is able to analyse the exported NanoString nCounter data and facilitates the user in performing a quality control. NACHO does this by visualising quality control metrics, expression of control genes, principal components and sample specific size factors in an interactive web application.

Calculate Overall Survival or Recurrence-Free Survival for breast cancer patients, using NHS Predict'. The time interval for the estimation can be set up to 15 years, with default at 10. Incremental therapy benefits are estimated for hormone therapy, chemotherapy, trastuzumab, and bisphosphonates. An additional function, suited for SCAN audits, features a more user-friendly version of the code, with fewer inputs, but necessitates the correct standardised inputs. This work is not affiliated with the development of NHS Predict and its underlying statistical model. Details on NHS Predict can be found at: <doi:10.1186/bcr2464>. The web version of NHS Predict': <https://breast.predict.nhs.uk/>. A small dataset of 50 fictional patient observations is provided for the purpose of running examples with the main two functions, and an additional dataset is provided for running example with the dedicated SCAN function.

This package implements methods introduced in Chen, Christensen, and Kankanala (2024) <doi:10.1093/restud/rdae025> for estimating and constructing uniform confidence bands for nonparametric structural functions using instrumental variables, including data-driven choice of tuning parameters. All methods in this package apply to nonparametric regression as a special case.

Th-U-Pb electron microprobe age dating of monazite, as originally described in <doi:10.1016/0009-2541(96)00024-1>.

Access the United States National Provider Identifier Registry API <https://npiregistry.cms.hhs.gov/api/>. Obtain and transform administrative data linked to a specific individual or organizational healthcare provider, or perform advanced searches based on provider name, location, type of service, credentials, and other attributes exposed by the API.

Create and manipulate numeric list ('nlist') objects. An nlist is an S3 list of uniquely named numeric objects. An numeric object is an integer or double vector, matrix or array. An nlists object is a S3 class list of nlist objects with the same names, dimensionalities and typeofs. Numeric list objects are of interest because they are the raw data inputs for analytic engines such as JAGS', STAN and TMB'. Numeric lists objects, which are useful for storing multiple realizations of of simulated data sets, can be converted to coda::mcmc and coda::mcmc.list objects.

Function and data sets in the book entitled "Nonlinear Time Series Analysis with R Applications" B.Guris (2020). The book will be published in Turkish and the original name of this book will be "R Uygulamali Dogrusal Olmayan Zaman Serileri Analizi". It is possible to perform nonlinearity tests, nonlinear unit root tests, nonlinear cointegration tests and estimate nonlinear error correction models by using the functions written in this package. The Momentum Threshold Autoregressive (MTAR), the Smooth Threshold Autoregressive (STAR) and the Self Exciting Threshold Autoregressive (SETAR) type unit root tests can be performed using the functions written. In addition, cointegration tests using the Momentum Threshold Autoregressive (MTAR), the Smooth Threshold Autoregressive (STAR) and the Self Exciting Threshold Autoregressive (SETAR) models can be applied. It is possible to estimate nonlinear error correction models. The Granger causality test performed using nonlinear models can also be applied.

In empirical studies, instrumental variable (IV) regression is the signature method to solve the endogeneity problem. If we enforce the exogeneity condition of the IV, it is likely that we end up with a large set of IVs without knowing which ones are good. Also, one could face the model uncertainty for structural equation, as large micro dataset is commonly available nowadays. This package uses adaptive group lasso and B-spline methods to select the nonparametric components of the IV function, with the linear function being a special case (naivereg). The package also incorporates two stage least squares estimator (2SLS), generalized method of moment (GMM), generalized empirical likelihood (GEL) methods post instrument selection, logistic-regression instrumental variables estimator (LIVE, for dummy endogenous variable problem), double-selection plus instrumental variable estimator (DS-IV) and double selection plus logistic regression instrumental variable estimator (DS-LIVE), where the double selection methods are useful for high-dimensional structural equation models. The naivereg is nonparametric version of ivregress in Stata with IV selection and high dimensional features. The package is based on the paper by Q. Fan and W. Zhong, "Nonparametric Additive Instrumental Variable Estimator: A Group Shrinkage Estimation Perspective" (2018), Journal of Business & Economic Statistics <doi:10.1080/07350015.2016.1180991> as well as a series of working papers led by the same authors.