

add1 {base}                                  R Documentation

_A_d_d _o_r _D_r_o_p _A_l_l _P_o_s_s_i_b_l_e _S_i_n_g_l_e _T_e_r_m_s _t_o _a _M_o_d_e_l

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

     Compute all the single terms in the `scope' argument
     that can be added to or dropped from the model, fit
     those models and compute a table of the changes in fit.

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

     add1(object, scope, ...)
     add1.default(object, scope, scale = 0, test=c("none", "Chisq"),
                  k = 2, trace = FALSE, ...)
     add1.lm(object, scope, scale = 0, test=c("none", "Chisq", "F"),
             x = NULL, k = 2, ...)
     add1.glm(object, scope, scale = 0, x = NULL, test=c("none", "Chisq"),
              k = 2, ...)

     drop1(object, scope, ...)
     drop1.default(object, scope, scale = 0, test=c("none", "Chisq"),
                   k = 2, trace = FALSE, ...)
     drop1.lm(object, scope, scale = 0, all.cols = TRUE,
              test=c("none", "Chisq", "F"),k = 2, ...)
     drop1.glm(object, scope, scale = 0, test=c("none", "Chisq"),
               k = 2, ...)

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

  object: a fitted models object.

   scope: a formula giving the terms to be considered for
          adding or dropping.

   scale: an estimate of the residual mean square to be used
          in computing Cp. Ignored if `0' or `NULL'.

    test: should the results include a test statistic rela-
          tive to the original model?  The F test is only
          appropriate for `lm' and `aov' models. The Chisq
          test can be an exact test (`lm' models with known
          scale) or a likelihood-ratio test depending on the
          method.

       k: the penalty constant in AIC/Cp.

   trace: if `TRUE', print out progress reports.

       x: a model matrix containing columns for the fitted
          model and all terms in the upper scope.  Useful if
          `add1' is to be called repeatedly.

all.cols: (Provided for compatibility with S.) Logical to
          specify whether all columns of the design matrix
          should be used. If `FALSE' then non-estimable
          columns are dropped, but the result is not usually
          statistically meaningful.

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

     For `drop' methods, a missing `scope' is taken to be
     all terms in the model. The hierarchy is respected when
     considering terms to be added or dropped: all main
     effects contained in a second-order interaction must
     remain, and so on.

     The methods for `lm' and `glm' are more efficient in
     that they do not recompute the model matrix and call
     the `fit' methods directly.

     The default output table gives AIC, defined as minus
     twice log likeliihood plus 2p where p is the rank of
     the model (the number of effective parameters).  This
     is only defined up to an additive constant (like log-
     likelhoods).  For linear Gaussian models with fixed
     scale, the constant is chosen to give Mallows' Cp,
     RSS/scale + 2p - n.  Where Cp is used, the column is
     labelled as Cp rather than AIC.

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

     An object of class `"anova"' summarizing the differ-
     ences in fit between the models.

_N_o_t_e_:

     These are not fully equivalent to the functions in S.
     There is no `keep' argument, and the methods used are
     not quite so computationally efficient.

     Their authors' definitions of Mallows' Cp and Akaike's
     AIC are used, not those of the authors of the models
     chapter of S.

_A_u_t_h_o_r_(_s_)_:

     B.D. Ripley

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

     `step', `aov', `lm', `extractAIC'.

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

     example(step)#-> swiss
     (alm1 <- add1(lm1, ~ I(Education^2) + .^2))

