Termplot               package:memisc               R Documentation

_P_r_o_d_u_c_e _a _T_e_r_m _P_l_o_t _L_a_t_t_i_c_e

_D_e_s_c_r_i_p_t_i_o_n:

     'Termplot' produces a lattice plot of termplots. Terms are not
     plotted individually, rather the terms in which a variable appears
     are summed and plotted.

_U_s_a_g_e:

       
     ## Default S3 method:
     Termplot(object,
             ...,
             variables=NULL,
             col.term = 2,
             lty.term = 1,
             lwd.term = 1.5,
             se = TRUE,
             col.se = "orange",
             lty.se = 2,
             lwd.se = 1,
             col.res = "gray",
             residuals = c("deviance","none","pearson","working"),
             cex = 1,
             pch = 1,
             jitter.resid=FALSE,
             smooth = TRUE,
             col.smth = "darkred",
             lty.smth = 2,
             lwd.smth = 1,
             span.smth = 2/3,
             aspect="fill",
             xlab=NULL,
             ylab=NULL,
             main=paste(deparse(object$call),collapse="\n"),
             models=c("rows","columns"),
             xrot = 0,
             layout=NULL)

_A_r_g_u_m_e_n_t_s:

  object: an model fit object, or a list of model objects

     ...: further model objects.

variables: a character vector giving the names of  independent
          variables; note that the combined effect of all terms
          containing the respective values will be plotted; if empty,
          the effect of each independent variable will be plotted.
          Currently, higher-order terms will be ignored.

col.term,lty.term,lwd.term: parameters for the graphical representation
          of the terms with the same meaning as in 'termplot'.

      se: a logical value; should standard error curves added to the
          plot?

col.se,lty.se,lwd.se: graphical parameters for the depiction of the
          standard error curves, see 'termplot'.

residuals: a character string, to select the type of residuals added to
          the plot.

col.res,cex,pch: graphical parameters for the depiction the residuals
          as in 'termplot'.

jitter.resid: a logical vector of at most length 2. Determines whether
          residuals should be jittered along the x-axis (first element)
          and along the y-axis. If this argument has length 1, its
          setting applies to both axes.

  smooth: a logical value; should a LOWESS smooth added to the plot?

span.smth: a numerical value, the span of the smoother.

col.smth,lty.smth,lwd.smth: graphical parameters for the smoother
          curve.

  aspect: aspect ratio of the plot(s), see 'xyplot'.

    xlab: label of the x axis, see 'xyplot'.

    ylab: label of the y axis, see 'xyplot'.

    main: main heading, see 'xyplot'.

  models: character; should models arranged in rows or columns?

    xrot: angle by which labels along the x-axis are rotated.

  layout: layout specification, see 'xyplot.'

_V_a_l_u_e:

     A trellis object.

_E_x_a_m_p_l_e_s:

     library(lattice)
     library(grid)


     lm0 <- lm(sr ~ pop15 + pop75,              data = LifeCycleSavings)
     lm1 <- lm(sr ~                 dpi + ddpi, data = LifeCycleSavings)
     lm2 <- lm(sr ~ pop15 + pop75 + dpi + ddpi, data = LifeCycleSavings)

     berkeley <- aggregate(Table(Admit,Freq)~.,data=UCBAdmissions)
     berk0 <- glm(cbind(Admitted,Rejected)~1,data=berkeley,family="binomial")
     berk1 <- glm(cbind(Admitted,Rejected)~Gender,data=berkeley,family="binomial")
     berk2 <- glm(cbind(Admitted,Rejected)~Gender+Dept,data=berkeley,family="binomial")

     Termplot(lm2)
     Termplot(berk2)
     Termplot(lm0,lm1,lm2)
     Termplot(berk0,berk1,berk2)

     Termplot(By(~Gender,glm(cbind(Admitted,Rejected)~Dept,family="binomial"),
                         data=berkeley))
     Termplot(By(~Dept,glm(cbind(Admitted,Rejected)~Gender,family="binomial"),
                         data=berkeley))

     require(splines)
     xyz <- data.frame(
       x = 1:100,
       z = factor(rep(LETTERS[1:4],25))
     )
     xyz <- within(xyz,
       y <- rnorm(100,sin(x/10)+x/50+as.numeric(z))
     )
     yxz.lin <- glm(y ~ x + z, data=xyz)
     yxz.bs <- glm(y ~ bs(x,6) + z, data=xyz)
     yxz.ns <- glm(y ~ ns(x,6) + z, data=xyz)
     yxz.poly <- glm(y ~ poly(x,6) + z, data=xyz)
     yxz.sincos <- glm(y ~ sin(x/10) + cos(x/10) + x + z, data=xyz)

     # Terms containing
     # the same variable are not plotted
     # individually but their combined effect is plotted
     #
     Termplot(yxz.lin,yxz.bs,yxz.ns,yxz.poly,yxz.sincos,models="columns",
       span.smth=1/3)

     Termplot(yxz.lin,yxz.bs,yxz.ns,yxz.poly,yxz.sincos,variables="x",
       span.smth=1/3)


       

