fit.normexp              package:limma              R Documentation

_F_i_t _N_o_r_m_a_l+_E_x_p _C_o_n_v_o_l_u_t_i_o_n _M_o_d_e_l _t_o _O_b_s_e_r_v_e_d _I_n_t_e_n_s_i_t_i_e_s

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

     Fit normal+exponential convolution model to observed intensities.
     The normal part represents the background and the exponential
     represents the signal intensities. This function is called by
     'backgroundCorrect' and is not normally called directly by the
     user.

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

     fit.normexp(foreground,background)
     fit.normexp0(foreground,background=NULL,background.matrix=NULL,trace=0,beta.start=NULL)

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

foreground: numeric vector of foreground intensities

background: vector of background intensity values

background.matrix: option design matrix for regression on background
          values

   trace: integer value passed to 'optim'. If positive then tracing
          information on the progress of the optimization is given. 
          Higher values give more information.

beta.start: optional numeric vector giving starting values for the
          regression coefficients

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

     These functions use maximum likelihood estimation to fit a model
     to the foreground and background intensities. The model is a
     $normal(mu,sigma^2)+exponential(alpha)$ convolution model for the
     foreground intensities in which $mu$ is a linear function of the
     background. 'fit.normexp' uses a BFGS modified Newton-Raphson
     iteration maxmize the likelihood while 'fit.normexp0' uses the
     Nelder-Mead simplex algorithm, a derivative free method.
     'fit.normexp' assumes that $mu$ is a constant plus the observed
     background. 'fit.normexp0' allows $mu$ to depend linearly on any
     set of covariates.

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

     A list containing the components 

    beta: numeric vector of estimated regression coefficients

   sigma: numeric scalar giving estimated value of $sigma$

   alpha: numeric scalar giving estimated value of $alpha$

m2loglik: numeric scalar giving minus twice the log-likelihood

convergence: integer code indicating successful convergence or
          otherwise of the optimization. See 'optim'.

_A_u_t_h_o_r(_s):

     Jeremy Silver and Gordon Smyth

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

     'signal.normexp', 'optim'

     An overview of normalization and background correction functions
     is given in '4.Normalization'.

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

     f <- c(2,3,1,10,3,20,5,6)
     b <- c(2,2,2,2,2,2,2,2)
     out <- fit.normexp(f,b)

