cells                package:Fahrmeir                R Documentation

_C_e_l_l_u_l_a_r _D_i_f_f_e_r_e_n_t_i_a_t_i_o_n

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

     The effect of two agents of immuno-activating ability that may
     induce cell differentiation was investigated.

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

     data(cells)

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

     A data frame with 16 observations on the following 3 variables.

     _y number of cells differentiating

     _T_N_F dose of TNF, U/ml

     _I_F_N dose of IFN, U/ml

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

     The effect of two agents of immuno-activating ability that may
     induce cell differentiation was investigated.  As response
     variable the number of cells that exhibited markers after exposure
     was recorded. It is of interest if the agents TNF (tumor necrosis
     factor) and IFN (interferon) stimulate cell differentiation
     independently, or if there is a synergetic effect.  200 cells were
     examined at each dose combination.

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

     Kjetil Halvorsen

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

     Ludwig Fahrmeir, Gerhard Tutz (1994):  Multivariate Statistical
     Modelling Based on Generalized Linear Models. Springer Series in
     Statistics.  Springer Verlag. New-York  Berlin Heidelberg

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

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

     str(cells)
     cells.poisson <- glm(y~TNF+IFN+TNF:IFN, data=cells, 
                         family=poisson)
     summary(cells.poisson)
     confint(cells.poisson)
     # Now we follow the book, example 2.6, page 51:
     # there seems to be overdispersion?
     cells.quasi <- glm(y~TNF+IFN+TNF:IFN, data=cells, 
                         family=quasipoisson)
     summary(cells.quasi)
     anova(cells.quasi)
     confint(cells.quasi)
     # We follow the book, example 2.7, page 56:
     with(cells, tapply(y, factor(TNF), function(x) c(mean(x), var(x))))
     # which might indicate the use of a negative binomial model

