R语言实战：统计与数据分析应用教程

需积分: 0 15 浏览量更新于2024-07-25 收藏 12MB PDF 举报

"《统计与数据分析：R语言应用实例》（Statistics and Data with R）是由耶西夫·科恩（Yosef Cohen）和杰里迈亚·Y·科恩（Jeremiah Y. Cohen）合著的一本专著，两位作者分别来自美国明尼苏达大学和范德比尔特大学，纳什维尔校区。该书于2008年由约翰威利与 Sons有限公司首次出版，ISBN号为978-0-470-75805-0。本书的核心内容是将统计学原理和数据处理技能通过丰富的实例呈现，强调了实践导向的学习方法。书中涵盖了广泛的统计知识，包括但不限于描述性统计、推断统计、概率论、假设检验、回归分析、时间序列分析以及在R语言中的实际操作。R语言作为一种开源的统计计算和图形软件，是现代数据分析的重要工具，书中详细介绍了如何利用R进行数据清洗、探索性分析、模型构建和结果可视化等步骤。作为一本应用于实践的教材，作者提供了大量实例来帮助读者理解复杂的统计概念，并展示了如何解决实际问题。每个例子都伴随着详细的代码解释，使读者能够逐步掌握R语言的基本语法和高级功能。此外，书中还涉及了如何处理缺失值、异常值，以及如何选择合适的统计方法来解决特定问题。版权方面，根据1988年英国《版权法》等相关法律规定，未经许可，不得以任何形式复制、存储或传输书中的内容，除非符合法律规定的例外情况。作者的署名权得到了充分尊重，读者可以在约翰威利与 Sons的官方网站上获取更多关于全球编辑办公室、客户服务以及如何申请使用版权材料的信息。综上，《统计与数据分析：R语言应用实例》是一本实用且深入的教程，适合对统计学感兴趣的学生、研究人员、数据分析师以及R语言初学者，它不仅教授理论知识，更注重通过实际操作提升读者的数据分析能力。"

Preface xvii

and ﬁrst-year graduate students in the sciences. The remaining Chapters, in a second

course. The book is accompanied by a solution manual which includes solutions to

most of the exercises. The solution manual is available through the book’s site.

• How should I study and use this book?

To study the book, read through each chapter. Each example provides enough

code to enable you to reproduce the results (statistical analysis and graphics) exactly

as shown. Scripts are explained in detail. Read through the explanation and try to

produce the results, as shown. This way, you ﬁrst see how a particular analysis is done

and how tables and graphics may be used to summarize it. Then when you study the

script, you will know what it does. Because of the adherence to the WYSIWYG

principle, some of the early examples might strike you as particularly complicated.

Be patient; as you progress, there are repetitions and you will soon learn the common

R idioms.

If you are familiar with both R and basic statistics, you may use the book as a

reference. Say you wish to run a simple logistic regression. You know how to do it

with R, but you wish to plot the regression on a probability scale. Then go to Chapter

16 and ﬂip through the pages until you hit Figure 16.5 in Example 16.5. Next, refer

to the example’s script and modify it to ﬁt your needs. There are so many R examples

and scripts that ﬂipping through the pages you are likely to ﬁnd one that is close to

what you need.

We include two indices, an R index and a general index. The R index is organized

such that functions are listed as ﬁrst entry and their arguments as sub-entries. The

page references point to applications of the functions by the various examples.

• Classical vs. Bayesian statistics

This topic addresses developments in statics in the last few decades. In recent

years and with advances in numerical computations, a trend among natural (and

other) scientists of moving away from classical statistics (where large data are needed

and no prior knowledge about it is assumed) to the so-called Bayesian statistics (where

prior knowledge about the data contributes to its analysis) has become fashionable.

Without getting into the sticky details, we make no judgment about the eﬃcacy of

one as opposed to the other approach. We prescribe to the following:

• With large data sets, statistical analyses using these two approaches often reach

the same conclusions. Developments in bootstrap methods make “small” data sets

“large”.

• One can hardly appreciate advances in Bayesian statistics without knowledge of

classical statistics.

• Certain situations require applications of Bayesian statistics (in particular when

real-time analysis is imperative). Others, do not.

• The analyses we present are suitable for both approaches and in most cases should

give identical results. Therefore, we pursue the classical statistics approach.

• A word about typography, data and scripts

We use monospaced characters to isolate our code work with R. Monospace lines

that begin with > indicate code as it is typed in an R session. Monospace lines that

Basic R

Here, we learn how to work with R. R is rooted in S, a statistical computing and

data visualization language originated at the Bell Laboratories (see Chambers and

Hastie, 1992). The S interpreter was written in the C programming language. R pro-

vides a wide variety of statistical and graphical techniques in a consistent computing

environment. It is also an interpreted programming environment. That is, as soon as

you enter a statement, or a group of statements and hit the Enter key, R computes

and responds with output when necessary. Furthermore, your whole session is kept

in memory until you exit R. There are a number of excellent introductions to and

tutorials for R (Becker et al., 1988; Chambers, 1998; Chambers and Hastie, 1992;

Dalgaard, 2002; Fox, 2002). Venables and Ripley (2000) was written by experts and

one can hardly improve upon it (see also The R Development Core Team, 2006b). An

excellent exposition of S and thereby of R, is given in Venables and Ripley (1994).

With minor diﬀerences, books, tutorials and applications written with S are rele-

vant in R. A good entry point to the vast resources on R is its website, http://

r-project.org.

R is not a panacea. If you deal with tremendously large data sets, with mil-

lions of observations, many variables and numerous related data tables, you will

need to rely on advanced database management systems. A particularly good one

is the open source PostgreSQL (http://www.postgresql.org). R deals eﬀectively—

and elegantly—with data that include several hundreds of thousands of observations

and scores of variables. The bulkier your system is in memory and hard disk space,

the larger the data set that R can handle. With all these caveats, R is a most ﬂexible

system for dealing with data manipulation and analysis. One often spends a large

amount of time preparing data for analysis and exploring statistical models. R makes

such tasks easy. All of this at no cost, for R is free!

This chapter introduces R. Learning R might seem like a daunting task, partic-

ularly if you do not have prior experience with programming. Steep as it may be,

the learning curve will yield much dividend later. You should develop tolerance

Statistics and Data with R: An applied approach through examples Y. Cohen and J.Y. Cohen

©2008 John Wiley & Sons, Ltd. ISBN: 978-0-470-75805

4 Basic R

ambiguity. Things might seem obscure at ﬁrst. Be patient and continue. With plenty

of repetitions and examples, the fog will clear. Being an introductory chapter, we are

occasionally compelled to make simple and sweeping statements about what R does

or does not do. As you become familiar with R, you will realize that some of our

statements may be slightly (but not grossly) inaccurate. In addition to R, we discuss

some abstract concepts such as classes and objects. The motivation for this seemingly

unrelated material is this: When you start with R, many of the commands (calls to

functions) and the results they produce may mystify you. With knowledge of the gen-

eral concepts introduced here, your R sessions will be clearer. To get the most out

of this chapter, follow this strategy: Read the whole chapter lightly and do whatever

exercises and examples you ﬁnd easy. Do not get stuck in a particular issue. Later, as

we introduce R code, revisit the relevant topics in this chapter for further clariﬁcation.

It makes little sense to read this chapter without actually implementing the code

and examples in a working R session. Some of the material in this chapter is based on

the excellent tutorial by Venables et al. (2003). Except for starting R, the instructions

and examples apply to all operating systems that R is implemented on. System speciﬁc

instructions refer to the Windows operating systems.

1.1 Preliminaries

In this section we learn how to work with R. If you have not yet done so, go to

http://r-project.org and download and install R. Next, start R. If you installed R

properly, click on its icon on the desktop. Otherwise, click on Start | Programs | R | R

x.y.z where x.y.z refers to the version number you installed. If all else fails, ﬁnd where

R is installed (usually in the Program Files directory). Then go to the bin directory

and click on the Rgui icon. Once you start R, you will receive a welcome statement

that ends with a prompt (usually >).

1.1.1 An R session

An R session consists of starting R, working with it and then quitting. You quit R

by typing q() or (in Windows) by selecting File | Exit.

R displays the prompt at the

beginning of the line as a cue for you. This is where you interact with the system

(you do not type the prompt). The vast majority of our work with R will consist of

executing functions. For now, we will say that a function is a small, single-purpose

executable program that R can be instructed to run. A function is identiﬁed by a

name followed by parentheses. Often the parentheses are separated by words. These

are called arguments. We say that we execute a function when we type the function’s

name, with the parentheses (and possibly arguments) and then hit the Enter key.

Let us go through a simple session. This will give you a feel for R. At any point,

if you wish to quit, type q() to end the session. To quit, you can also use the File |

Exit menu. At the prompt, type (recall that you do not type the prompt):

help.start()

As a rule, while working with R, we avoid menus and graphical interface and stick with

the command line interface.

剩余602页未读，继续阅读

jiangdmdr

粉丝: 58
资源: 767

R语言实战：统计与数据分析应用教程

Statistics and Data Analysis for Financial Engineering

Practical Statistics for Data Scientists: 50 Essential Concepts

Applied spatial data analysis with R

Statistics.and.Data.with.R.An.applied.approach.through.examples

Biostatistics with R_ An Introduction to Statistics Through Biological Data

Missing and Modified Data in Nonparametric Estimation with R

Introductory Statistics with R.pdf

Multivariate Statistics with R,2009

Graphing Data with R(O'Reilly,2015)

Doing Bayesian Data Analysis: A Tutorial with R and BUGS

最新资源