FScomputeEnvelope           package:Rfwdmv           R Documentation

_C_o_m_p_u_t_e _a _C_o_n_f_i_d_e_n_c_e _E_n_v_e_l_o_p_e

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

     Computes a confidence envelope for the Random Starts forward plot.

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

     FScomputeEnvelope(n, p, m, probs = c(0.01, 0.5, 0.99), tolerance = 0.001)

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

       n: an integer value giving the number of observations in the
          data.

       p: an integer value giving the dimension of the data.

       m: an integer giving the size of the smallest subset in the
          forward search.

   probs: a numeric vector containing the probabilities used to compute
          the quantiles.

tolerance: a positive numeric value giving the convergence tolerance.

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

     a numeric matrix whose columns contain the quantiles of the given
     order statisics.  The row names contain the subset sizes and the
     columns names contain the quantiles specified in 'probs'.

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

     'fwdmvRandomStart',  'plot.fwdmvRandomStart'.

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

     X <- matrix(rnorm(1200), nrow = 200, ncol = 6)
     temp <- fwdmvRandomStart(X)

     # the envelope is computed by FScomputeEnvelope #
     plot(temp, envelope = FALSE)
     envelope <- FScomputeEnvelope(200, 6, 12)
     domain <- as.numeric(row.names(envelope))
     correction <- 200*6*(domain-2)/((200-1)*(domain-6-1))
     envelope <- sqrt(sweep(envelope, 1, correction, "*"))
     matlines(domain, envelope, lty = 2, col = 1)

