

gee(gee)                                     R Documentation

_F_u_n_c_t_i_o_n _t_o _s_o_l_v_e _a _G_e_n_e_r_a_l_i_z_e_d _E_s_t_i_m_a_t_i_o_n _E_q_u_a_t_i_o_n _M_o_d_e_l

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

     Produces an object of class "gee" which is a General-
     ized Estimation Equation fit of the data.

_U_s_a_g_e_:

     gee(formula, id,
         data, subset, na.action,
         R=NA, b=NA,
         tol=0.001, maxiter=as.integer(25),
         family = gaussian, corstr="independence",
         Mv=1, silent=T, contrasts=NULL,
         scale.fix = F, scale.value = 1, v4.4compat=F)

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

 formula: a formula expression as for other regression mod-
          els, of the form response ~ predictors. See the
          documentation of lm and formula for details.

      id: a vector which identifies the clusters.  The
          length of `id' should be the same as the number of
          observations.  Data are assumed to be sorted so
          that observations on a cluster are contiguous rows
          for all entities in the formula.

    data: an optional data frame in which to interpret the
          variables occurring in the `formula', along with
          the `id' and `n' variables.

  subset: expression saying which subset of the rows of the
          data should  be  used in the fit.  This can be a
          logical vector (which is replicated to have length
          equal to the number of observations), or a numeric
          vector indicating which observation numbers are to
          be included, or a  character  vector of the row
          names to be included.  All observations are
          included by default.

na.action: a function to filter missing data.  For `gee'
          only `na.omit' should be used here.

       R: a square matrix of dimension maximum cluster size
          containing the user specified correlation.  This
          is only appropriate if `corstr="fixed"'.

       b: an initial estimate for the parameters.

     tol: the tolerance used in the fitting algorithm.

 maxiter: the maximum number of iterations.

  family: a `family' object: a list of functions and expres-
          sions for defining link and variance functions.
          Families supported in `gee' are `gaussian`, `bino-
          mial', `poisson', `Gamma', and `quasi'; see the
          `glm' and `family' documentation.  Some links are
          not currently available: `1/mu^2' and `sqrt' have
          not been hard-coded in the cgee engine at present.
          The inverse gaussian variance function is not
          available.  All combinations of remaining func-
          tions can be obtained either by family selection
          or by the use of `quasi'.  Future releases will
          allow S-coded functions for link, variance and
          correlation function specification.

  corstr: a character specifying the Correlation structure.
          The following are permitted: `"independence"'
          `"fixed"' `"stat_M_dep"' `"non_stat_M_dep"'
          `"exchangeable"' `"AR-M"' `"unstructured"'

      Mv: When the corstr is `"stat_M_dep"',
          `"non_stat_M_dep"', or `"AR-M"' then `Mv' must be
          specified.

  silent: a logical variable controlling whether parameter
          estimates at each iteration are printed.

contrasts: a list giving contrasts for some or all of the
          factors appearing in the model formula.  The ele-
          ments of the list should have the same name as the
          variable and should be either a contrast matrix
          (specifically, any full-rank matrix with as many
          rows as there are levels in the factor), or else a
          function to compute such a matrix given the number
          of levels.

scale.fix: a logical variable; if true, the scale parameter
          is fixed at the value of `scale.value'

scale.value: numeric variable giving the value to which the
          scale parameter should be fixed; used only if
          `scale.fix == T'.

v4.4compat: logical variable requesting compatibility of
          correlation parameter estimates with previous ver-
          sions; the current version revises to be more
          faithful to the Liang and Zeger (1986) proposals
          (compatible with the Groemping SAS macro, version
          2.03)

_D_e_t_a_i_l_s_:

     Though input data need not be sorted by the variable
     named `"id"', the program will interpret physically
     contiguous records possessing the same value of `id' as
     members of the same cluster.  Thus it is possible to
     use the following vector as an `id' vector to discrimi-
     nate 4 clusters of size 4:
     `c(0,0,0,0,1,1,1,1,0,0,0,0,1,1,1,1)'.

_V_a_l_u_e_:

     An object of class `"gee"' representing the fit.

_S_i_d_e _E_f_f_e_c_t_s_:

     Offsets must be specified in the model formula, as in
     `glm()'.

_N_O_T_E_:

     This is version 4.8 of this user documentation file,
     revised 98/01/27.  The assistance of Dr B Ripley is
     gratefully acknowledged.

_R_e_f_e_r_e_n_c_e_s_:

     Liang, K.Y. and Zeger, S.L. (1986).  Longitudinal data
     analysis using generalized linear models.  Biometrika
     73 13-22.

     Zeger, S.L. and Liang, K.Y. (1986).  Longitudinal data
     analysis for discrete and continuous outcomes.  Biomet-
     rics 42 121-130.

_S_e_e _A_l_s_o_:

     `glm', `lm', `formula'.

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

     gee(y ~ x,id)   # Gaussian model with independent correlation structure

     gee(d ~ x1+x2,id,link="logit",family=binomial,corstr="exchangeable")

     gee(d ~ x1+x2,id=id,family=binomial(link=(probit)),
         corstr="stat_M_dep",M=1)

