loess.control             package:stats             R Documentation

_S_e_t _P_a_r_a_m_e_t_e_r_s _f_o_r _L_o_e_s_s

_D_e_s_c_r_i_p_t_i_o_n:

     Set control parameters for 'loess' fits.

_U_s_a_g_e:

     loess.control(surface = c("interpolate", "direct"),
                   statistics = c("approximate", "exact"),
                   trace.hat = c("exact", "approximate"),
                   cell = 0.2, iterations = 4, ...)

_A_r_g_u_m_e_n_t_s:

 surface: should be fitted surface be computed exactly or via
          interpolation from a kd tree?

statistics: should the statistics be computed exactly or approximately?
          Exact computation can be very slow.

trace.hat: should the trace of the smoother matrix be computed exactly
          or approximately? It is recommended to use the approximation
          for more than about 1000 data points.

    cell: if interpolation is used this controls the accuracy of the
          approximation via the maximum number of points in a  cell in
          the kd tree. Cells with more than 'floor(n*span*cell)' points
          are subdivided.

iterations: the number of iterations used in robust fitting.

     ...: further arguments which are ignored.

_V_a_l_u_e:

     A list with components 

 surface: 

statistics: 

trace.hat: 

    cell: 

iterations: 

     with meanings as explained under 'Arguments'.

_S_e_e _A_l_s_o:

     'loess'