GRI                  package:qualV                  R Documentation

_A _G_e_o_m_e_t_r_i_c _R_e_l_i_a_b_i_l_i_t_y _I_n_d_e_x _a_s _S_u_g_g_e_s_t_e_d _b_y _L_e_g_g_e_t_t & _W_i_l_l_i_a_m_s

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

     Given a set of predictions and a corresponding set of
     observations, the geometric validation index is a reliability
     index for the predictions.

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

     GRI(o, p)

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

       o: vector of observed values

       p: vector of corresponding predicted values

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

     One possible interpretation of 'GRI' is that the simulation is
     accurate within a multiplicative factor 'GRI', i.e. the observed
     values fall between 1/GRI and GRI times of the corresponding
     predicted values. Values close to one indicate a good match.

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

     GRI: geometric reliability index

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

     Leggett, L. R. and Williams, L. R. (1981) A reliability index for
     models. Ecological Modelling, 13, 303-312.

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

     'MAE', 'MSE', 'MAPE', 'EF'

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

     # a constructed example
     x <- seq(0, 2*pi, 0.1)
     y <- 5 + sin(x)           # a process
     o <- y + rnorm(x, sd=0.2) # observation with random error
     p <- y + 0.1              # simulation with systematic bias

     plot(x, o); lines(x, p)
     GRI(o, p)

     # observed and measured data with non-matching time intervals
     data(phyto)
     obsb <- na.omit(obs[match(sim$t, obs$t), ])
     simb <- sim[na.omit(match(obs$t, sim$t)), ]
     GRI(obsb$y, simb$y)

