kootenay             package:robustbase             R Documentation

(_M_o_d_i_f_i_e_d) _W_a_t_e_r_f_l_o_w _M_e_a_s_u_r_e_m_e_n_t_s _o_f _K_o_o_t_e_n_a_y _R_i_v_e_r _i_n _L_i_b_b_y _a_n_d _N_e_w_g_a_t_e

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

     The original data set is the waterflow in January of the Kootenay
     river, measured at two locations, namely, Libby (Montana) and
     Newgate (British Columbia) for 13 consecutive years, 1931-1943.

     The data set is of mostly interest because it has been used as
     example in innumerous didactical situations about robust
     regression. To this end, one number (in observation 4) has been
     modified from the original data from originally 44.9 to 15.7
     (here).

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

     data(kootenay)

_F_o_r_m_a_t:

     A data frame with 13 observations on the following 2 variables.

     '_L_i_b_b_y' a numeric vector

     '_N_e_w_g_a_t_e' a numeric vector

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

     Other modified versions of the data sets are also used in
     different places, see the examples below.

_S_o_u_r_c_e:

     Original Data, p.58f of Ezekiel and Fox (1959), _Methods of
     Correlation and Regression Analysis_. Wiley, N.Y.

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

     Hampel, F., Ronchetti, E., Rousseeuw, P. and Stahel, W.  (1986)
     _Robust Statistics: The Approach Based on Influence Functions_;
     Wiley, N.Y.

     Rousseeuw, P.~J. and Leroy, A.~M. (1987) _Robust Regression &
     Outlier Detection_, Wiley, N.~Y.

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

     data(kootenay)
     plot(kootenay, main = "'kootenay' data")
     points(kootenay[4,], col = 2, cex =2, pch = 3)

     abline(lm   (Newgate ~ Libby, data = kootenay), col = "pink")
     abline(lmrob(Newgate ~ Libby, data = kootenay), col = "blue")

     ## The original version of Ezekiel & Fox:
     kootenay0 <- kootenay
     kootenay0[4, "Newgate"] <- 44.9
     plot(kootenay0, main = "'kootenay0': the original data")
     abline(lm   (Newgate ~ Libby, data = kootenay0), col = "pink")
     abline(lmrob(Newgate ~ Libby, data = kootenay0), col = "blue")

     ## The version with "milder" outlier -- Hampel et al., p.310
     kootenay2 <- kootenay0
     kootenay2[4, "Libby"] <- 20.0 # instead of 77.6
     plot(kootenay2, main = "The 'kootenay2' data",
          xlim = range(kootenay[,"Libby"]))
     points(kootenay2[4,], col = 2, cex =2, pch = 3)
     abline(lm   (Newgate ~ Libby, data = kootenay2), col = "pink")
     abline(lmrob(Newgate ~ Libby, data = kootenay2), col = "blue")

