MASS: Support Functions and Datasets for Venables and ...

1 Package mass July 14, $Rev: 3600 $DependsR (>= ), grDevices, graphics, stats, utilsImportsmethodsSuggestslattice, nlme, nnet, survival, gamairDescriptionFunctions and Datasets to Support Venables and Ripley,``Modern Applied Statistics with S'' (4th edition, 2002).TitleSupport Functions and Datasets for Venables and Ripley's MASSLazyDatayesByteCompileyesLicenseGPL- 2 | Ripley [aut, cre, cph],Bill Venables [ctb],Douglas M. Bates [ctb],Kurt Hornik [trl] (partial port ca 1998),Albrecht Gebhardt [trl] (partial port ca 1998),David Firth [ctb]MaintainerBrian 10:19:17 UTC12 Rtopics documented:Rtopics documented:abbey ..5accdeaths ..5addterm ..6 Aids2 ..7 Animals ..8anorexia .. 10area .. 11bacteria .. 13bcv .. 14beav1 .. 15beav2 .. 16 Belgian-phones .. 17biopsy .. 18birthwt .. 19 Boston .. 20boxcox .. 21cabbages .. 22caith.

2 23 Cars93 .. 24cats .. 25cement .. 26chem .. 27con2tr .. 27confint- mass .. 29coop .. 30corresp .. 35cpus .. 36crabs .. 37 Cushings .. 38 DDT .. 38deaths .. 39denumerate .. 40drivers .. 41dropterm .. 42eagles .. 43epil .. 44eqscplot .. 46farms .. 47fgl .. 48fitdistr .. 49 Rtopics documented:3forbes .. 51fractions .. 51 GAGurine .. 52galaxies .. 55gehan .. 56genotype .. 57geyser .. 58gilgais .. 58ginv .. 61glmmPQL .. 62hills .. 65housing .. 65huber .. 67hubers .. 68immer .. 69 Insurance .. 70isoMDS .. 71kde2d .. 72lda .. 74ldahist .. 76leuk .. 79loglm .. 81logtrans .. 83lqs .. 85mammals .. 87mca .. 88mcycle .. 89 Melanoma .. 90menarche .. 90michelson .. 91minn38 .. 92motors .. 92muscle.

3 93mvrnorm .. 96newcomb .. 97nlschools .. 97npk .. 98npr1 .. 100 Null .. 100oats .. 1014 Rtopics documented:OME .. 102painters .. 106parcoord .. 107petrol .. 112polr .. 121qda .. 122quine .. 124 Rabbit .. 125rational .. 126renumerate .. 127rlm .. 130rnegbin .. 131road .. 132rotifer .. 133 Rubber .. 133sammon .. 134ships .. 135shoes .. 136shrimp .. 136shuttle .. 137 Sitka .. 137 Sitka89 .. 138 Skye .. 139snails .. 140SP500 .. 141stdres .. 141steam .. 142stepAIC .. 143stormer .. 145studres .. 148survey .. 151abbey5topo .. 153 Traffic .. 153truehist .. 154ucv .. 155 UScereal .. 156 UScrime .. 157VA .. 158waders .. 159whiteside .. 162wtloss .. 163 Index165abbeyDeterminations of Nickel ContentDescriptionA numeric vector of 31 determinations of nickel content (ppm) in a Canadian syenite Abbey (1988)Geostandards Newsletter12, , W.

4 N. and Ripley, B. D. (2002)Modern Applied Statistics with edition. Deaths in the US 1973-1978 DescriptionA regular time series giving the monthly totals of accidental deaths in the values for first six months of 1979 (p. 326) were7798 7406 8363 8460 9217 J. Brockwell and R. A. Davis (1991)Time Series: Theory and , New , W. N. and Ripley, B. D. (2002)Modern Applied Statistics with All One-Term Additions to a ModelDescriptionTry fitting all models that differ from the current model by adding a single term from those supplied,maintaining function is generic; there exist methods for classeslmandglmand the default method willwork for many other (object, ..)## Default S3 method:addterm(object, scope, scale = 0, test = c("none", "Chisq"),k = 2, sorted = FALSE, trace = FALSE, ..)## S3 method for class'lm'addterm(object, scope, scale = 0, test = c("none", "Chisq", "F"),k = 2, sorted = FALSE.)

5 ## S3 method for class'glm'addterm(object, scope, scale = 0, test = c("none", "Chisq", "F"),k = 2, sorted = FALSE, trace = FALSE, ..)ArgumentsobjectAn object fitted by some model -fitting formula specifying a maximal model which should include the current additional terms in the maximal model with all marginal terms in the originalmodel are in the definition of the AIC statistic for selecting the models, currently onlyforlm,aovandglmmodels. Specifyingscaleasserts that the residual standarderror or dispersion is the results include a test statistic relative to the original model ? The F testis only appropriate forlmandaovmodels, and perhaps for some over-dispersedglmmodels. The Chisq test can be an exact test (lmmodels with known scale)or a likelihood-ratio test depending on the multiple of the number of degrees of freedom used for the penalty. Onlyk=2gives the genuine AIC:k = log(n)is sometimes referred to as BIC or the results be sorted on the value of AIC?

6 TraceifTRUE additional information may be given on the fits as they are passed to or from other definition of AIC is only up to an additive constant: when appropriate (lmmodels with specifiedscale) the constant is taken to be that used in Mallows Cp statistic and the results are table of class"anova"containing at least columns for the change in degrees of freedom and AIC(or Cp) for the models. Some methods will give further information, for example sums of squares,deviances, log-likelihoods and test , W. N. and Ripley, B. D. (2002)Modern Applied Statistics with edition. Alsodropterm, <- aov(log(Days + ) ~ .^4, quine) <- aov(log(Days+ ) ~ 1, quine)addterm( , , test="F") <- glm(Freq ~ Infl*Type*Cont + Sat, family=poisson,data=housing)addterm( , ~. + Sat:(Infl+Type+Cont), test="Chisq") <- update( , . ~ . + Sat*(Infl+Type+Cont))addterm( , ~. + Sat:(Infl+Type+Cont)^2, test = "Chisq")Aids2 Australian AIDS Survival DataDescriptionData on patients diagnosed with AIDS in Australia before 1 July data frame contains 2843 rows and the following columns:stateGrouped state of origin:"NSW "includes ACT and"other"is WA, SA, NT and of (Julian) date of (Julian) date of death or end of "A"(alive) or"D"(dead) at end of transmission (years) at data set has been slightly jittered as a condition of its release, to ensure patient P.

7 J. Solomon and the Australian National Centre in HIV Epidemiology and Clinical , W. N. and Ripley, B. D. (2002)Modern Applied Statistics with edition. and Body Weights for 28 SpeciesDescriptionAverage brain and body weights for 28 species of land weight in weight in nameAnimalsavoided conflicts with a system datasetanimalsin S-PLUS and J. Rousseeuw and A. M. Leroy (1987)Robust Regression and Outlier , p. , W. N. and Ripley, B. D. (2002)Modern Applied Statistics with Data on Weight ChangeDescriptionTheanorexiadata frame has 72 rows and 3 columns. Weight change data for young femaleanorexia data frame contains the following columns:TreatFactor of three levels:"Cont"(control),"CBT"(Cognitive Behavioural treatment) and"FT"(family treatment).PrewtWeight of patient before study period, in of patient after study period, in , D. J., Daly, F.

8 , McConway, K., Lunn, D. and Ostrowski, E. eds (1993)A Handbook of SmallData & Hall, Data set 285 (p. 229)(Note that the original source mistakenly says that weights are in kg.)ReferencesVenables, W. N. and Ripley, B. D. (2002)Modern Applied Statistics with edition. Ratio Tests for Negative binomial GLMsDescriptionMethod function to perform sequential likelihood ratio tests for Negative binomial generalizedlinear ## S3 method for class'negbin'anova(object, .., test = "Chisq")ArgumentsobjectFitted model object of class"negbin", inheriting from classes"glm"and"lm",specifying a Negative binomial fitted GLM. Typically the output ()..Zero or more additional fitted model objects of class"negbin". They shouldform a nested sequence of models, but need not be specified in any to match thetestargument Ignored (with a warningif changed) if a sequence of two or more Negative binomial fitted model objectsis specified, but possibly used if only one object is function is a method for the generic functionanova()for class"negbin".

9 It can be invoked bycallinganova(x)for an objectxof the appropriate class, or directly by (x)regardless of the class of the only one fitted model object is specified, a sequential analysis of deviance table is given for thefitted model . Thethetaparameter is kept fixed. If more than one fitted model object is specifiedthey must all be of class"negbin"and likelihood ratio tests are done of each model within the this casethetais assumed to have been re-estimated for each , W. N. and Ripley, B. D. (2002)Modern Applied Statistics with edition. , , <- (Days ~ Eth*Age*Lrn*Sex, quine, link = log)m2 <- update(m1, . ~ . - Eth:Age:Lrn:Sex)anova(m2, m1)anova(m2)areaAdaptive Numerical IntegrationDescriptionIntegrate a function of one variable over a finite range using a recursive adaptive method. Thisfunction is mainly for demonstration (f, a, b, .., fa = f(a.))

10 , fb = f(b, ..),limit = 10, eps = 1e-05)ArgumentsfThe integrand as anSfunction object. The variable of integration must be thefirst limit of limit of arguments needed by the value at the lower value at the upper on the depth to which recursion is allowed to tolerance to control the method divides the interval in two and compares the values given by Simpson s rule and thetrapezium rule. If these are within eps of each other the Simpson s rule result is given, otherwisethe process is applied separately to each half of the interval and the results added integral fromatoboff(x).ReferencesVenables, W. N. and Ripley, B. D. (1994)Modern Applied Statistics with (sin, 0, pi) # integrate the sin function from 0 to of Bacteria after Drug TreatmentsDescriptionTests of the presence of the bacteriaH. influenzaein children with otitis media in the NorthernTerritory of data frame has 220 rows and the following columns:ypresence or absence: a factor with : a factor with compliance: a factor with : week of ID: a factor with levelsplacebo,druganddrug+, a re-coding A.