Rossiter-LiterateDat..
Contents
1. by the package but since Sweave is not interactive the analyst must specify it directly It s a simple trick for example 12 oO N o TE T g R g e ge J o Oo 0 e 8 e e j ee a e 65 70 75 80 85 Height feet Figure 1 Relation between girth and height 31 cherry trees lt lt fig TRUE gt gt require lattice provides the xyplot method for scatterplots data trees print with trees xyplot Volume Girth This behaviour is useful when collecting several lattice graphics for display in a matrix with the split and more arguments to print lt lt fig TRUE width 8 height 4 gt gt require lattice data trees pi lt with trees xyplot Volume Girth p2 lt with trees xyplot Volume Height print pl split c 1 1 2 1 more T print p2 split c 2 1 2 1 more F 3 3 Manipulating variables used in graphics In the previous example you might be tempted to add rm p1 p2 to remove the temporary variables This will cause an error because the chunk with graphics output is run twice silently once to produce the figure and once to 13 produce any printed output Hence data manipulations including deleting variables should be done either in separate chunks lt lt gt gt rm pi p2 3 4 R code formatting and comments If you are an experienced R programmer you probably do two things that are good programming practice e Formattin2. text editors Emacs ESS Tinn R you can directly send lines or chunks of code from the NoWeb source to a linked R console otherwise you have to work in the two environments separately Thus you have an interactive data analysis as you work but write it up in a document to be read by others Note The terms Weave and Tangle are from Knuth a reference to a poem by Sir Walter Scott Oh what a tangled web we weave when first we practise to deceive Marmion V1 17 Knuth s original literate programming system was called WEB so he decided to use Weave for the process of making the readable document and Tangle for the process of making the executable code So now you know We now give a tutorial example and then get into some of the details and complications 8 NoWeb http www cs tufts edu nr noweb 2 Tutorial We will do a small literate data analysis on one of R s example datasets trees 1 Examine the dataset structure 2 Summarize the variables 3 Graph the relation between them 4 Build a linear model to predict tree volume from tree girth and height All of this is accompanied by our commentary this is where we explain literately we hope what we are doing why and what conclusions we draw 2 1 First version Task 1 Create the BTEX master file named test tex open it in the text editor and set up the ATEX document This is the usual document skeleton naming t
3. 10 13 summary 1m 10 Sweave 3 6 9 Sweave package 8 trees dataset 4 19 A Source code A 1 ETpX master file documentclass 11pt article usepackage Sweave title Modelling tree volume author D Luo date 11 November 2011 begin document maketitle Here we use the verb trees dataset supplied with R to illustrate a simple data analysis 1 describing the variables and cases 2 investigating the inter relation between variables 3 modelling tree volume as a function of tree height and or tree girth input test1 tex end document A 2 First version of Sweave source par First load the dataset examine its structure and summarize the variables par lt lt gt gt data trees str trees summary trees endinput It is always good practice to end TX source to be included in a master document with the endinput macro 20 A 3 Second version of Sweave source SweaveOpts prefix string graph test eps false setkeys Gin width 0 6 textwidth First load the dataset examine its structure and summarize the variables lt lt gt gt data trees str trees summary trees par Second look at the pairwise scatterplots of the three variables lt lt fig T width 7 height 7 gt gt pairs trees pch 20 cex 1 2 par Comment There appears to be a very strong relation between girth and volume this seems slightly non linear parabolic T
4. 13 is to also write Encapsulated Postscript EPS figures Second it is good practice to create a subdirectory to hold the graphics there will typically be a lot of them and they clutter up the main project directory Task 9 Create a subdirectory named graph Task 10 Add the following line at the beginning of your NoWeb source i e the Rnw file Sweave0pts prefix string graph test eps false This says to put figures in the graph subdirectory relative to the working directory prefix the names with test and not write EPS files Graphs will be generated with the names test 001 pdf test 002 pdf etc and these names will be used in the generated source file with the includegraphics PTEX command Another issue with graphics is their size on the page in the generated PDF document By default the Sweave R package specifies 0 8 times the current text width this leaves a space of 0 1 times the text width at each side If the figure doesn t have much complexity you might want it narrower if a wide figure landscape orientation you might want it wider At any point in the document you can change it with the Gin graphics inches option of the setkeys TFX command defined by Sweave sty e g setkeys Gin width 0 6 textwidth Note It s most convenient to define the width in terms of textwidth as shown above however direct specification of width is also permitted The code chunk to produce a figure looks like th
5. 50 60 70 10 20 3 40 50 60 70 Comment There appears to be a very strong relation between girth and volume this seems slightly non linear parabolic The relation between height and volume is also positive but much weaker Height and girth are very weakly related this suggests that the trees have different morphologies Figure 4 Second output with a graph page 2 of 2 26 8 0 2 4 16 18 a 10 20 3 40 50 60 70 Comment There appears to be a very strong relation between girth and volume this seems slightly non linear parabolic The relation between height and volume is also positive but much weaker Height and girth are very weakly related this suggests that the trees have different morphologies Third model the tree volume by a full model with the two possible predictors include the interaction R gt m lt lm Volume Girth Height data trees R gt summary m Call lm formula Volume Girth Height data trees Residuals Min 1Q Median 3Q Max 6 582 1 067 0 303 1 564 4 665 Coefficients Estimate Std Error t value Pr gt t Intercept 69 3963 23 8358 2 91 0 00713 Girth 5 8558 1 9213 3 05 0 00511 Height 1 2971 0 3098 4 19 0 00027 Girth Height 0 1347 0 0244 5 52 7 5e 06 Residual standard error 2 71 on 27 degrees of freedom Multiple R squared 0 976 Adjusted R squared 0 973 F statistic 359 on 3 and 27 DF p value lt 2e 16 The succe
6. 831 3 F Leisch Sweave Dynamic generation of statistical reports using literate data analysis Compstat 2002 Proceedings in Computational Statistics pages 575 580 2002 3 Friedrich Leisch Sweave user s manual R version 2 7 1 2008 URL http wiw stat uni muenchen de leisch Sweave Sweave manual pdf 12 17 Edzer J Pebesma Multivariable geostatistics in S the gstat package Computers amp Geosciences 30 7 683 691 2004 12 R Development Core Team R A Language and Environment for Sta tistical Computing R Foundation for Statistical Computing Vienna Austria 2011 URL http www R project org ISBN 3 900051 07 0 3 D G Rossiter Introduction to the R Project for Statistical Computing for use at ITC University of Twente Faculty ITC 3 85 edition Nov 2010 URL http www itc nl personal rossiter teach R RIntro_ITC pdf 16 Deepayan Sarkar Lattice R News 2 2 19 23 2002 12 Deepayan Sarkar Lattice multivariate data visualization with R Springer 2008 ISBN 9780387759685 pbk 0387759689 pbk 9780387759692 e ISBN 0387759697 e ISBN URL http 1mdvr r forge r project org 12 18 Index of R Concepts cor test 13 dev off 11 graphics package 11 gstat package 11 jpeg 11 lattice package 11 more lattice graphics argument 12 pdf 11 pdffonts 14 png 11 print package lattice 12 print 11 R WinEdt package 2 source 3 10 split lattice graphics argument 12 Stangle 3
7. DF file and may be used by itself e g for a journal article or thesis These will have names like test 001 pdf according to the prefix string argument to the SweaveDpts macro see 2 2 However you may want a different formatting for a production graphic To do this within a code chunk open a graphics device with the pdf jpeg or png functions write code to produce the graph and close the graphics device with the dev off function For example lt lt gt gt pdf file graph scatterGirthHeight pdf width 5 height 5 title Figure 1 bg lightgray fg darkred plot trees Girth trees Height pch 20 cex 1 5 xlab Height feet ylab Girth inches dev off Note there is no fig TRUE in the chunk header because we produce the graphic by hand rather than automatically Also notice the many options that can be given the function that opens the graphics device here pdf This produces the nice graphic shown in Figure 1 3 2 Lattice graphics R has several graphics systems the example above uses base graphics from the graphics package which is always loaded with R Another very sophisticated graphics system is provided by the lattice package 8 9 this is used by other packages such as the gstat geostatis tics package 5 Lattice graphics do not produce output directly instead they return a lattice object which can be printed with the generic print method In interactive R this is done automatically
8. Technical Note Literate Data Analysis Contents D G Rossiter April 13 2011 1 Overview 3 2 Tutorial 5 21 First yersin o ik eutr wa Se eb eS ee Pee eee ee 5 22 Addn graphics s s 4 de a a s ek as d oe i 8 23 In line calculations 2 orana aara ee et 10 2 4 Writing an R source code file 11 3 Details 12 a Prodiction graphics sss ae riana anepi e eee et 12 Soe JLattieewrapligs i ac ca wow bog a ee A ee eee 12 3 3 Manipulating variables used in graphics 13 3 4 R code formatting and comments 14 3 5 Hiding code from the reader 15 3 6 Hiding output from the reader 15 4 Learning to use the tools 16 A DIRX 2c aoe hi Ree eG ee bee ee ee eee eS 16 Ae WG g a ea een ees Ba ee ee eee ees oP eg 16 AS EMACS a scant e e ker ee ee dee Pe eb ee ee ee a 16 AA weave sa oco ccoo de ob shoa aoa Radau ws we me 17 References 18 Version 1 1 Copyright 2011 D G Rossiter All rights reserved Repro duction and dissemination of the work as a whole not parts freely per mitted if this original copyright notice is included Sale or placement on a web site where payment must be made to access this document is strictly prohibited To adapt or translate please contact the author http www itc nl personal rossiter Index of R concepts A Source code A 1 TATRX master file 2 A 2 First version of Sweave source A 3 Second
9. acs Learning Emacs is an investment in a lifetime of programming productivity but not an overnight busi ness e Microsoft Windows only WinEdt and the R WinEdt R package to communicate with it e Microsoft Windows only Tinn R The flow is as follows Thttp www r project org 2nttp www stat uni muenchen de leisch Sweave 3http www latex project org 4 http www gnu org software emacs 5 http ess r project org 6 http www winedt com 1 http www sciviews org Tinn R 1 You create a source document in a text editor with extension Rnw a so called NoWeb file this source document includes IATRX markup your own text and chunks of executable R code using the NoWeb syntax explained below to show which parts of the source are exe cutable code 2 You run this through R with the R function Sweave S language Weave this produces a IATFX file extension tex which includes your original IAT X markup and text with the output from R which may include graphics 3 You process the MTFX file with PDFIATRX to produce a PDF docu ment 4 Optional You run the original source through R with the R function Stangle S language Tangle to produce an R source code file with the same name and extension R this can be executed with the R function source As you create your source document you can also execute lines or chunks of code in the R console to see their effect From some
10. eave to convert it to a ATEX source named test1 tex So we need to include this TeX file which will be produced by Sweave in the master file Task 4 Use the input MITEX macro to include the Sweave output in the master file after the introductory text input test1 tex After these steps my master file is as shown in A 1 Task 5 Create a new source file named test1 Rnw and open it in the text editor Note The Rnw extension is used for NoWeb source files Task 6 Write the code and commentary to load the example dataset e For this first example you just need to know one NoWeb syntax a code chunk is written between lt lt gt gt and these must be the only text on their respective lines of NoWeb source Anything between these is considered R code and will be formatted executed and the output written to the ATpx source file lt lt gt gt R code here Anything not in a code chunk is regular ATRX source this is where you write comments and explanations My code and commentary is shown in A 2 Task 7 Sweave this source file within R with the Sweave function this creates a ATFX file with the same name but extension tex gt Sweave test1 Rnw Writing to file testi tex Processing code chunks You can now run LaTeX on testi tex The resulting file should look like B 1 Notice the TFX environments provided by the Sweave IATFX package Schunk for any S language chunk
11. es Girth num 8 3 8 6 8 8 10 5 10 7 10 8 11 11 11 1 11 2 Height num 70 65 63 72 81 83 66 75 80 75 Volume num 10 3 10 3 10 2 16 4 18 8 19 7 15 6 18 2 22 6 19 9 R gt summary trees Girth Min SUB dst Qu 11 Median 12 Mean 213 3rd Qu 15 Max 20 Figure 2 First output 24 Modelling tree volume D Luo 11 November 2011 Here we use the trees dataset supplied with R to illustrate a simple simple data analysis 1 describing the variables and cases 2 investigating the inter relation between variables 3 modelling tree volume as a function of tree height and or tree girth First load the dataset examine its structure and summarize the vari ables R gt data trees R gt str trees data frame 31 obs of 3 variables Girth num 8 3 8 6 8 8 10 5 10 7 10 8 11 11 11 1 11 2 Height num 70 65 63 72 81 83 66 75 80 75 Volume num 10 3 10 3 10 2 16 4 18 8 19 7 15 6 18 2 22 6 19 9 R gt summary trees Girth Height Volume Min 8 Min 263 Min 10 dst Qu 11 ist Qu 72 ist Qu 19 Median 12 Median Median 24 Mean 213 Mean 76 Mean 730 3rd Qu 15 3rd Qu 3rd Qu 37 Max 20 Max 87 Max Ha Second look at the pairwise scatterplots of the three variables R gt pairs trees pch 20 cex 1 2 Figure 3 Second output with a graph page 1 of 2 25 8 10 2 4 16 18 a Height 10 20 30 40
12. f you don t know how to use them An excellent starting point is the IATRX Wikibook This explains instal lation simple and advanced usage and tricks It includes an Absolute Be ginners section Of course the IATFX project home page is the definitive portal The R environment for statistical computing home page is the entry point for information downloads and documentation I have written an introduction to R for ITC 7 10 lists some learning resources The most useful for beginners is Appendix A A sample session of the Introduction to R from the R Project This will give you some familiarity with the style of R sessions and more importantly some instant feedback on what actually happens Don t worry if you don t understand everything this is just to give you a feel for how R works and what it can do For individual commands it is always best to look at its help topic Many other introductions R have been written both as formal textbooks and on line documents see the Documents link in the table of contents of the R home page If you choose to use Emacs you face a steep learning curve but end up witha programming and text editing environment of unequalled power and speed The reference manual at the GNU Emacs home page is comprehensive and systematic but slow going The same group produces an Emacs Tour which shows some of the capabilities Probably the best way to get started is t
13. g your code for readability for example adding line breaks e Addding R comments introduced with the character to explain your R code You can do these in your Sweave source but by default they will not appear in your final document however they will appear in any R code generated with the Stangle function see 2 4 This is because Sweave runs the source code through the R parser and itself formats the result as printed output So what to do Comments are not so necessary in literate programming because you can explain things in the text so in most cases the default behaviour is fine If you really want formatted code or comments in your document you can use the keep source TRUE option for a single code chunk For example suppose we want to keep our comment and formatting to ex plain a correlation test using the cor test function lt lt keep source T gt gt a non parametric rank correlation between the two predictors cor test trees Girth trees Height method spearman Here we separated the lines for easier reading and explain with a comment that the test is non parametric Of course this could have been explained in the text not as a code comment This will appear in the document as 14 gt now do the correlation gt cor test trees Girth trees Height method spearman Spearman s rank correlation rho data trees Girth and trees Height S 2773 4 p value 0 01306 alterna
14. he document class loading packages etc A minimal skeleton is documentclass 11pt article begin document LaTeX macros and text go here end document There is usually a title author and date documentclass 11pt article title Modelling tree volume author D Luo date 11 November 2011 begin document maketitle LaTeX macros and text go here end document To have properly formatted Sweaved text you must load that IATFX package in the document preamble Task 2 Add the following macros before begin document i e in the preamble usepackage Sweave Task 3 Write the introductory text in the document section of BTEX master file i e within the document environment This should be your description to your reader of the purpose of this data analysis Here is my text Here we use the verb trees dataset supplied with R to illustrate a simple simple data analysis 1 describing the variables and cases 2 investigating the inter relation between variables 3 modelling tree volume as a function of tree height and or tree girth I find it easiest to have a master file with the TFX headers package decla rations options etc and then use the input or include macro if each section should start on a new page to include one or more files which contain the results of R computation and my comments on them Below we will create a NoWeb source file named test1 Rnw and process it with Sw
15. he relation between height and volume is also positive but much weaker Height and girth are very weakly related this suggests that the trees have different morphologies endinput 21 A 4 Third version of Sweave source SweaveOpts prefix string graph test eps false setkeys Gin width 0 6 textwidth First load the dataset examine its structure and summarize the variables lt lt gt gt data trees str trees summary trees par Second look at the pairwise scatterplots of the three variables lt lt fig T width 7 height 7 gt gt pairs trees pch 20 cex 1 2 par Comment There appears to be a very strong relation between girth and volume this seems slightly non linear parabolic The relation between height and volume is also positive but much weaker Height and girth are very weakly related this suggests that the trees have different morphologies par Third model the tree volume by a full model with the two possible predictors include the interaction lt lt gt gt note is used to specify an interaction effect m lt 1lm Volume Girth Height data trees summary m The success is quite good as measured by the adjusted R 2 Sexpr round summary m adj r squared 100 1 endinput Notice the calculation in the Sexpr macro the result is multipled by 100 to express it as a percentage and then rounded to one decimal place 22 B Intermed
16. iate files B 1 First version of Sweave generated BTpX source par First load the dataset examine its structure and summarize the variables par begin Schunk begin Sinput R gt data trees R gt str trees end Sinput begin Soutput data frame 31 obs of 3 variables Girth num 8 3 8 6 8 8 10 5 10 7 10 8 11 11 11 1 11 2 Height num 70 65 63 72 81 83 66 75 80 75 Volume num 10 3 10 3 10 2 16 4 18 8 19 7 15 6 18 2 22 6 19 9 end Soutput begin Sinput R gt summary trees end Sinput begin Soutput Girth Height Volume Min 8 3 Min 63 Min 710 2 ist Qu 11 1 ist Qu 72 ist Qu 19 4 Median 12 9 Median 76 Median 24 2 Mean 713 2 Mean 76 Mean 730 2 3rd Qu 15 2 3rd Qu 80 3rd Qu 37 3 Max 20 6 Max 87 Max 77 0 end Soutput end Schunk endinput Note the automatically generated Schunk Sinput and Soutput ATEX envi ronments interpreted by the Sweave 4TRX package C Output These are Figures 2 3 and 5 23 Modelling tree volume D Luo 11 November 2011 Here we use the trees dataset supplied with R to illustrate a simple data analysis 1 describing the variables and cases 2 investigating the inter relation between variables 3 modelling tree volume as a function of tree height and or tree girth First load the dataset examine its structure and summarize the vari ables R gt data trees R gt str trees data frame 31 obs of 3 variabl
17. is lt lt fig T gt gt R code to produce graphics e g plot hist You can also specify the dimensions of the PDF graphic in the code chunk header e g lt lt fig T width 10 height 5 gt gt R code to produce graphics e g plot hist These dimensions are inches default is 6 x6 Fonts are scaled to look good for the graphic printed on standard A4 paper so specifying a larger size results in smaller fonts relative to the graphic elements With this preparation we can add a graph to our test document Task 11 1 Add code to the NoWeb source to draw a graph 2 Also display the graph interactively to check the graph is what you want and to interpret it 3 Add some interpretative text to the NoWeb source explaining the graph 91 2 54 cm 72 points exactly 4 Sweave this source file within R with the Sweave function 5 TEXify the master file My interepretation was Comment There appears to be a very strong relation between girth and volume this seems slightly non linear parabolic The relation between height and volume is also positive but much weaker Height and girth are very weakly related this suggests that the trees have different morphologies Now when we Sweave the source this commentary is given right after the figure The reader can see the figure and the analyst s interpretation My revised NoWeb source with graphics commands and some comments is sh
18. o follow the tutorial built in to Emacs This is accessed by using the help system and then pressing the t for tutorial key Unfortunately different platforms and even different keyboard mappings have different ways to access the help system 10 http en wikibooks org wiki LaTeX 11 http www latex project org 12 http www r project org 13 http cran r project org doc manuals R intro pdf p proy g P 14 http www gnu org software emacs Manuals 15 http www gnu org software emacs tour 16 e Under X11 or Mac OS X terminal press the lt fl gt key If you start Emacs without a file name the opening screen explains how to access the help system Emacs has many useful extensions which may be installed by default or you may have to install them For editing MTFX source the AUCTRX extension can be used For communicating with R and running R within the Emacs editor the solution is the ESS Emacs Speaks Statistics extension 4 4 Sweave The Sweave manual 4 has full explanation and useful examples 16 http www gnu org software auctex 17 http ess r project org 17 References 1 Donald Ervin Knuth Literate programming Center for the Study of Language and Information 1992 ISBN 0937073814 cloth 0937073806 paper 3 Leslie Lamport LaTeX a document preparation system user s guide and reference manual Addison Wesley Pub Co 1994 ISBN 0201529
19. own in A 3 After Sweaving this source gt Sweave test1 Rnw we get the PDF file shown in Figure 3 2 3 In line calculations Sweave is also able to write calculated numbers right into the text For example you might want to comment on the success of a model with some thing like The adjusted R of the model is quite high 0 86 But how do you know the figure You could compute it interactively in R and then cut and paste but that is error prone and would have to be repeated if you change the model or dataset Far better is to use the Sexpr ATEX macro provided by the Sweave ATRX style Most R expressions that produce a sin gle number can be arguments to this macro the results of the R calculation are then written to the IATEX source when the source file is Sweaved For example the ATX source text Sexpr round 2 pi 360 5 10 will produce 0 01745 in the document In practice you compute interactively in R see what works and then add the relevant output to your in line text in the NoWeb source From some text editors Emacs ESS Tinn R you can directly send lines or chunks of code from the NoWeb source to a linked R console otherwise you have to work in the two environments separately Task 12 Compute a linear model of tree volume modelled as an interaction between height and girth and report its goodness of fit in line with the Sexpr TFX macro Explain the processing steps in the text and interp
20. produced from the NoWeb source Sinput for formatted S language input and Soutput for formatted R output Note In general you never have to look at this file it is generated automat ically by Sweave and included in your PDFETRX output with the input or include macros We show it so that you can see what Sweave does Task 8 T Xify the document run PDFIATRX to produce the PDF file which will be named test pdf The output should look like Figure 2 2 2 Adding graphics Sweave can produce graphical output in two ways 1 The author specifies ig TRUE in the code chunk header and writes the usual graphics commands in the code chunk a figure is automatically generated named stored on your computer and incorporated in the PDF via the BTEX includegraphics macro 2 The author explicitly opens a graphics device e g a PDF file and writes to it with the usual graphics commands within a code chunk The second option is only needed if you want to generate a figure formatted for publication see 3 1 for details There are a few details that make this process go smoothly The first is to use the IAT X like SweaveOpts macro in the NoWeb source Rnw file before any R code that produces graphics to specify two things e The location and the prefix of the file name of automatically produced graphics files the default is the current directory and source file name e That we only need PDF figures the default for R lt 2
21. ret the result e Here I examine the model summary interactively and decide to report the goodness of fit as an adjusted R this is given by the adj r squared field of the model summary given by summary 1m My revised NoWeb source is shown in A 4 After Sweaving this source by running the R command gt Sweave test1 Rnw to produce file test1 tex and TpXifying the master file we get the PDF file shown in Figure 5 2 4 Writing an R source code file You may want the R code as a separate file for inclusion in an automatic process or as source for further experimentation This is the function of the Tangle procedure You do this interactively at the R prompt using the Stangle function Task 13 Tangle the final NoWeb source to produce R code e Recall the source code is in file test1 Rnw So at the R prompt gt Stangle test1 Rnw The result is shown in 8D This source can now be run in R with the source function gt source test1 R This would run all the analysis and produce all the graphics but not the document 11 3 Details The Sweave manual 4 has full explanation of the many options useful examples and some common tricks Here we just list a few that may catch the unwary 3 1 Production graphics The graphics produced automatically with the lt lt fig TRUE gt gt code chunk header are included in your PDF document and stored on your system Each graphic is a separate P
22. ss is quite good as measured by the adjusted R 97 3 Figure 5 Third output with a graph and in line calculation page 2 of 2 Page 1 is the same as Figure 3 27 D Generated R source code HHEHHHHHHHHHHEHHEHHEHHEHHHAHHHHHHHEHHEHHHHHEHHHHE EH chunk number 1 HHEHHHHHHHHHHEHHEHHEHHAHHAHHHHHHHEHHEHHEHHEHHHHE EH line 5 test3 Rnw data trees str trees summary trees HHEHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHEHHHHHEHHE HEH chunk number 2 HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHEHHRHHEHHH HHH line 12 test3 Rnw pairs trees pch 20 cex 1 2 HEHE HHA AE EE AREA chunk number 3 HEHEHE HHH HARARE ARR aaa line 24 test3 Rnw note is used to specify an interaction effect m lt 1lm Volume Girth Height data trees summary m Note the automatically generated comments marked with the character Also note that the line in the source NoWeb file is given so we can easily find which code chunk produced with R code 28
23. tive hypothesis true rho is not equal to 0 sample estimates rho 0 44084 Note how the formatting and comment are preserved along with R s first line and continuation prompts You can also include this as one of the Sweave options for an entire source file recall from 2 2 that the ATFX like SweaveOpts macro is placed at the start of the NoWeb source Rnw file For example SweaveOpts keep source TRUE remainder of source file 3 5 Hiding code from the reader You may want to execute some code that is irrelevant to readers for example changing to a directory on your system that will not be on their systems You can hide code with the echo FALSE tag lt lt echo F gt gt setwd Users Goliath projects secret notell Any output will not be hidden 3 6 Hiding output from the reader You may want to hide some output probably because it is too long or verbose but you want to show the reader what you did You can hide output with the results hide tag You can show all the PDF fonts on your system as follows lt lt results hide gt gt str pdfFonts This will appear in the document as You can show all the PDF fonts on your system as follows gt str pdfFonts without any of the voluminous output produced by pdffonts 15 4 Learning to use the tools 4 1 BTpx 4 2 R 4 3 Emacs We ve explained the interaction between the various tools here we list some resources to get you started i
24. version of Sweave source A 4 Third version of Sweave source Intermediate files B 1 First version of Sweave generated IATFX source Output Generated R source code 23 28 1 Overview In 1992 Donald Knuth published a book with the title Literate Program ming 1 showing the advantages of and techniques for writing computer programs to be read and understood by humans as well as executed by a digital computer This technical note advocates the same approach for data analysis the executable computer code here in the R environment is an integral part of a document that explains what the analyst did why and what was discovered The advantages of this approach are several 1 every processing step is transparent 2 anyone else can repeat the analysis if they are given access to the same data 3 analysis can easily be expanded or adapted 4 the results of the analysis are generated with the document so they are by definition synchronized 5 the analyst s motivations and interpretations are in the exact place where the results of the analysis are presented The tools we use are Data processing The R environment for statistical computing 6 Literate programming Sweave 3 Text processing TEX 2 Text editor There are several good choices e Emacs with the AUCTRX extension for working with IATFX doc uments and the ESS Emacs Speaks Statistics extension for running R under Em
Download Pdf Manuals
Related Search
Rossiter LiterateDat..