R语言实战：探索与分析

下载需积分: 9 | PDF格式 | 2.54MB | 更新于2024-07-22 | 20 浏览量 | 举报

"R in Action" 是一本由Robert I. Kabacoff编写的关于R语言实战应用的书籍，旨在帮助读者深入理解和应用R语言进行数据分析。书中包含了大量的编程实例，覆盖了从数据预处理到高级统计分析的多个方面。本书分为三个主要部分，详细介绍了R语言在数据分析中的应用： 1. 第一部分：准备数据 - 第1章介绍R语言，包括R的历史、安装和基本使用，以及R Studio等开发环境的设置。 - 第2章讲解如何创建数据集，包括数据输入、数据结构（如向量、列表、数据框和因子）以及数据导入导出的技巧。 - 第3章和第4章涉及基础和进阶的数据管理，如数据清洗、子集选取、排序、合并与重塑数据，以及处理缺失值和异常值的方法。 2. 第二部分：基础统计和图形 - 第5章介绍了统计学的基础概念，如描述性统计、中心趋势度量、变异度量以及假设检验。 - 第6章讲解如何使用R绘制基本图形，包括散点图、直方图、箱线图等，并讨论了如何定制图形的外观和进行统计图形的解释。 3. 第三部分：中级统计和图形 - 第7章探讨了多元线性回归，包括模型构建、参数估计、诊断和预测。 - 第8章介绍了方差分析(ANOVA)，用于比较多个组间的差异，同时涵盖了单因素和多因素的ANOVA。 - 第9章讲解了重采样统计和自助法（bootstrapping），这是估计统计量的不确定性和评估模型稳健性的非参数方法。 - 第10章关注功效分析，帮助读者理解实验设计时如何确定样本大小以确保足够的检测力。 - 第11章可能涉及更高级的统计主题，但具体内容未给出。通过这些章节，读者将能够熟练掌握R语言进行数据探索、建模和可视化的基本技巧，以及如何运用这些工具解决实际数据分析问题。此外，书中提供的在线论坛链接（http://www.manning-sandbox.com/forum.jspa?forumID=5782）为读者提供了一个讨论和获取帮助的平台，有助于深化学习和交流。

16 Robert I. Kabacoff / R in Action Last saved: 8/10/2009

©Manning Publications Co. Please post comments or corrections to the Author Online forum:

http://www.manning-sandbox.com/forum.jspa?forumID=578

Let’s put it all together by looking at the example in listing 1.3

Listing 1.3 Using various input and output in an R session

source("myfile1") 1

sink(“myoutput”, append=TRUE, split=TRUE)

pdf(“mygraphs.pdf”)

source(“myfile2”) 2

sink()

dev.off()

source(“myfile3”) 3

In the code above, R statements from

myfile1 are submitted to the current session #1

and the results appear on the screen. When the statements from

myfile2 are submitted

#2, results appear on the screen, the text output is appended to the file

myoutput, and the

graphic output is saved to

mygraphs.pdf. Finally the statements from myfile3 are

submitted #3 and the results appear on screen.

R provides quite a bit of flexibility and control over where input comes from and where it

goes. In section 1.5 we will see how to run a program in batch mode.

1.4 Packages

R comes with extensive capabilities right out of the box. However, some of its most exciting

features are available as optional modules that you can download and install. There are over

1800 user contributed modules called packages that you can download from http://cran.r-

project.org/web/packages. They provide a tremendous range of new capabilities, from the

analysis of geostatistical data to protein mass spectra processing to the analysis of

educational tests! We will use many of these optional packages in this book.

1.4.1 What are packages?

Packages are collections of R functions, data, and compiled code in a well-defined format.

The directory where packages are stored on your computer is called the library. The function

.libPaths() will show you where your library is located, while the function library()

will show you what packages you have saved in your library.

As we've said, R comes with a standard set of packages, while others are available for

download and installation. Once installed, they have to be loaded into the session in order to

be used. The command

search() will tell you which packages are loaded and ready to use.

1.4.2 Installing a package

There are a number of R functions that let you manipulate packages. To install a package for

the first time, use the

install.packages() command. For example,

install.packages() without options will bring up a list of CRAN mirror sites. Once you

select a site, you will be presented with a list of all available packages. Selecting one will

Licensed to Alison Tyler <pedbro@gmail.com>

Download at Boykma.Com

Last saved: 8/10/2009 Robert I. Kabacoff / R in Action 17

©Manning Publications Co. Please post comments or corrections to the Author Online forum:

http://www.manning-sandbox.com/forum.jspa?forumID=578

download and install it. If you know what package you want to install, you can do so directly

by providing it as an argument to the function. For example, the

gclus package contains

functions for creating enhanced scatter plots. You can download and install the package with

the command

install.packages("glus").

You only need to install a package once. However, like any software, packages are often

updated by their authors. Use the command

update.packages() to update any packages

that you have installed. To see details on your packages, you can use the

installed.packages() command. It will list the packages you have, along with their

version numbers, dependencies, and other information.

1.4.3 Loading a package

Installing a package downloads it from a CRAN mirror site and places it in your library. To

actually use it in an R session, you need to load the package using the library() command.

For example, to use the packaged

gclus, issue the command library(gclus). Of course,

you must have installed a package before you can load it. You will have to load a package

once in each session you want to use it. However, you can customize your start-up

environment to automatically load the packages you use most often. Customizing your start-

up is covered in appendix x.

1.4.4 Learning about a package

When you load a package, a new set of functions and datasets become available. Small

illustrative datasets are provided along with sample code, allowing you to try out the new

functionalities. The help system contains a description of each function (along with

examples), and information on each dataset included. Entering

help(package="name")

will provide a brief description of the package named and an index of the functions and

datasets included. Using

help()with any of these function or dataset names will provide

further details. The same information can be downloaded as a PDF manual from CRAN.

Common mistakes in R programming

There are some common mistakes made frequently by both beginning and experienced R

programmers. If your program generates an error be sure the check for the following:

Using the wrong case. help(), Help(), and HELP() are three different functions (only

the first will work).

Forgetting to use quote marks when they are needed. install.packages("gclus")

will work, while install.packages(gclus) will generate an error.

Forgetting to include the parentheses in a function call. help() rather than help. Even

if there are no options, you still need the ().

Licensed to Alison Tyler <pedbro@gmail.com>

Download at Boykma.Com

18 Robert I. Kabacoff / R in Action Last saved: 8/10/2009

©Manning Publications Co. Please post comments or corrections to the Author Online forum:

http://www.manning-sandbox.com/forum.jspa?forumID=578

Using the \ in a path name on Windows. R sees the backslash character as an escape

character. setwd("c:\mydata") will generate an error. Use setwd("c:/mydata")

or setwd("c:\\mydata") instead.

Using a function from a package that is not loaded. The function order.clusters() is

contained in the gclus package. If you try to use it before loading the package, you will

get an error.

The error message in R can be cryptic, but if you are careful to follow the points above,

you should avoid seeing many of them.

In the sidebar above, these should be bullet points, with the text

of the first sentence bolded for each. Bullets start with "Using the

wrong case" and end with "Using a function".

1.5 Batch Processing

Most of the time, you will be running R interactively, entering commands at the command

prompt and seeing the results of each statement as it is processed. Occasionally, you may

want to run an R program in a repeated, standard, and possibly unattended fashion. For

example, you may need to generate the same report once a month. You can write your

program in R and run it in batch mode.

How you run R in batch mode depends on your operating system. On Linux or MacOS X

systems, you can use the following command in a terminal window:

R CMD BATCH options infile outfile

where

infile is the name of the file containing R code to executed, outfile is name

of the file receiving the output and

options lists options that control execution. By

convention, the infile is given extension

.R and the outfile is given extension .Rout.

For Windows, use

"C:\Program Files\R\R-2.9.0\bin\R.exe" CMD BATCH [CA]

--vanilla --slave "c:\my projects\myscript.R"

adjusting the paths to match the location of your R.exe binary and your script file. For

more details on how to invoke R, including command line options, see an "Introduction to R"

from CRAN (http://cran.r-project.org

1.6 Using output as input - Reusing results

One of the most useful design features of R is that the output of analyses can easily be

saved and used as input to additional analyses. Let's walk through an example. If you don't

understand the statistics involved, don't worry. We are focusing on the general principle

here.

Licensed to Alison Tyler <pedbro@gmail.com>

Download at Boykma.Com

Last saved: 8/10/2009 Robert I. Kabacoff / R in Action 19

http://www.manning-sandbox.com/forum.jspa?forumID=578

This following code will run a simple linear regression of miles per gallon (mpg) on car

weight (wt) using the dataset mtcars. Results are sent to the screen. Nothing is saved.

lm(mpg~wt, data=mtcars)

This time, the same regression is performed but the results are saved under the name fit.

fit <- lm(mpg~wt, data=mtcars)

No output is sent to the screen. However, you now can manipulate the results.

The assignment has actually created a list called "fit" that contains a wide range of

information from the analysis (including the predicted values, residuals, regression

coefficients, and more). Typing

summary(fit) provides details of the analysis, while

plot(fit) produces diagnostic plots. You can generate and save influence statistics with

cook<-cooks.distance(fit). plot(cook) will graph these influence statistics. To

predict miles per gallon from car weight in a new set of data use

predict(fit,

mynewdata)

To see what a function returns, look at the value section of the online help for that

function. Here we would look at

help(lm). This will tell you what is saved when you assign

the results of that function to a name.

1.7 Working through an example

We will finish this chapter with an example that ties many of these ideas together. Here is

the task:

1. Open up the general help and look at the Introduction to R section.

2. Install the vcd package (a package for visualizing categorical data that we will be

using in future chapters).

3. List the functions and datasets available in this package.

4. Load the package and read the description of the dataset Arthritis.

5. Print out the Arthritis data set (entering the name of an object will list it).

6. Run the example that comes with the Arthritis dataset. Don't worry if you don't

understand the results. It basically shows that arthritis patients receiving treatment

improved much more than patients receiving a placebo.

7. Quit

The code required is provided in listing 1.4, with a sample of the results displayed in figure

1.6.

Listing 1.4 Working with a new package

help.start() # look at introduction and preliminaries

install.packages("vcd")

Licensed to Alison Tyler <pedbro@gmail.com>

Download at Boykma.Com

剩余117页未读，继续阅读

身份认证购VIP最低享 7 折!

30元优惠券

wangzh1109

粉丝: 1

R语言实战：探索与分析

Ajax in Action英文原版电子书

提升英语能力与实践：JSF in Action 英文原版指南

权威指南：iBATIS in Action 英文原版PDF

掌握Apache Spark：《Spark in Action》原版PDF分享

Struts in Action 中英文对照

英文原版-The Intelligent Enterprise in the Era of Big Data 1st Edition

AndroidinAction3rdEdition.pdf 英文原版

英文原版-Hidden Markov Models for Time Series 2nd Edition

SMT常用术语中英文对照.pdf

R语言速成：153分钟掌握基础

最新资源