$23.00

# Statistical Computing in C++ and R

## Preview

## Book Description

With the advancement of statistical methodology inextricably linked to the use of computers, new methodological ideas must be translated into usable code and then numerically evaluated relative to competing procedures. In response to this, **Statistical Computing in C++ and R** concentrates on the writing of code rather than the development and study of numerical algorithms per se. The book discusses code development in C++ and R and the use of these symbiotic languages in unison. It emphasizes that each offers distinct features that, when used in tandem, can take code writing beyond what can be obtained from either language alone.

The text begins with some basics of object-oriented languages, followed by a "boot-camp" on the use of C++ and R. The authors then discuss code development for the solution of specific computational problems that are relevant to statistics including optimization, numerical linear algebra, and random number generation. Later chapters introduce abstract data structures (ADTs) and parallel computing concepts. The appendices cover R and UNIX Shell programming.

**Features**

- Includes numerous student exercises ranging from elementary to challenging
- Integrates both C++ and R for the solution of statistical computing problems
- Uses C++ code in R and R functions in C++ programs
- Provides downloadable programs, available from the authors’ website

The translation of a mathematical problem into its computational analog (or analogs) is a skill that must be learned, like any other, by actively solving relevant problems. The text reveals the basic principles of algorithmic thinking essential to the modern statistician as well as the fundamental skill of communicating with a computer through the use of the computer languages C++ and R. The book lays the foundation for original code development in a research environment.

## Table of Contents

**Introduction **Programming paradigms

Object-oriented programming

What lies ahead

**Computer representation of numbers**

Introduction

Storage in C++

Integers

Floating-point representation

Errors

Computing a sample variance

Storage in R

Exercises

**A sketch of C++**

Introduction

Variables and scope

Arithmetic and logical operators

Control structures

Using arrays and pointers

Functions

Classes, objects and methods

Miscellaneous topics

Matrix and vector classes

.Input, output and templates

.Function templates

.Exercises

**Generation of pseudo-random numbers**

Introduction

Congruential methods

Lehmer type generators in C++

An FMclass

Other generation methods

Nonuniform generation

Generating random normals

Generating random numbers in R

Using the R Standalone Math Library

.Exercises

**Programming in R**

Introduction

File input and output

Classes, methods and namespaces

Writing R functions

Avoiding loops in R

An example

Using C/C++ code in R

Exercises

**Creating classes and methods in R**

Introduction

Creating a new class

Generic methods

An example

Exercises

**Numerical linear algebra**

Introduction

Solving linear equations

Eigenvalues and eigenvectors

Singular value decomposition

Least squares

The Template Numerical Toolkit

Exercises

**Numerical optimization**

Introduction

Function objects

Golden section

Newton’s method

Maximum likelihood

Random search

Exercises

**Abstract data structures**

Introduction

ADT dictionary

ADT priority queue

ADT ordered set

Pointer arithmetic, iterators and templates

Exercises

**Data structures in C++**

Introduction

Container basics

Vector and deque

The C++ list container

Queues

The map and set containers

Algorithm basics

Exercises

Parallel computing in C++ and R

Introduction

OpenMP

Basic MPI commands for C++

Parallel processing in R

Parallel random number generation

Exercises

A An introduction to Unix

A.Getting around and finding things

A.Seeing what’s there

A.Creating and destroying things

A.Things that are running and how to stop them

B An introduction to R

B.R as a calculator

B.R as a graphics engine

B.R for statistical analysis

C C++ library extensions (TR)

C.Pseudo-random numbers

C.Hash tables

C.Tuples

D The Matrix and Vector classes

E The ranGen class

References

Index

## Reviews

"…the first treatment of parallel programming in R that I have seen in a book. The text is replete with code examples and there are numerous end-of-chapter exercises."

—International Statistical Review, 2013