# Plot individual MNREAD fitted curves as estimated by a nonlinear mixed-effect (NLME) modeling. This function uses the NLME model created from `nlmeModel` to plot individual curves and Critical Print Size (CPS). ```r nlmeCurve(nlme.model, displayCPS = TRUE, CPScriterion = NULL) ``` ## Arguments - `nlme.model`: The object returned by `nlmeModel` - `displayCPS`: Optional argument to display the CPS on the individuals curves. Default is TRUE. If set to FALSE, the CPS won't be plotted. - `CPScriterion`: Optional argument to specify a criterion for CPS estimation. The default criterion value is '90 of MRS'. This criterion can vary from 75 to 95 of MRS and should only be modified for specific purposes, as discussed in Cheung et al. 2008 ## Returns The function returns a plot of reading speed (in log words/min) as a function of print size (in logMAR). If displayCPS is not specified, the Critical Print Size will be marked as an inverted triangle. ## Notes Print size shown on the plot(s) have been corrected for non-standard testing viewing distance. For more details on the nlme fit, see:\ Cheung SH, Kallie CS, Legge GE, Cheong AM. Nonlinear mixed-effects modeling of MNREAD data. Invest Ophthalmol Vis Sci. 2008;49:828–835. doi: 10.1167/iovs.07-0555. ## Examples ```r # inspect the structure of the dataframe head(data_low_vision, 10) #------ # restrict dataset to one MNREAD test per subject (regular polarity only) data_regular <- data_low_vision %>% filter (polarity == "regular") # run the NLME model for data grouped by subject nlme_model <- nlmeModel(data_regular, ps, vd, rt, err, subject) #------ # plot MNREAD curves and CPS with a default CPS criterion of '90 of MRS' nlmeCurve(nlme_model) # plot MNREAD curves without the CPS for a default CPS criterion of '90 of MRS' nlmeCurve(nlme_model, FALSE) # plot MNREAD curves and CPS with a specific CPS criterion of '80 of MRS' nlmeCurve(nlme_model, TRUE, 0.8) #------ # Once created, the NLME curve can be further customized using ggplot2 # plot the NLME curve my_plot <- nlmeCurve(nlme_model) # display my.plot print(my_plot) # modify my.plot my_new_plot <- my_plot + # overwrites the raw data points geom_point(data = nlme_model[[1]], aes(x=correct_ps, y = rs), size = 4) + # changes the colors of the curve and raw data (effective only for nested designs) scale_color_brewer(palette="Set1") + # changes the colors of the CPS diamond (effective only for nested designs) scale_fill_brewer(palette="Set1") + # modifies the aspect of the x-axis scale_x_continuous(breaks = seq (-0.5,2.5,0.4)) # display my.new.plot print(my_new_plot) #------ # For very large datasets, it can be usefull to plot only selected facets to inspect individual fit # To do so, one needs to restrict the dataframe called in each of the three layers of the plot # list of subject names to keep subjects_to_keep <- paste ("s", 1:4, sep = "") # first filter the original data points (data called in the first layer) my_plot$data <- my_plot$data %>% filter(subject %in% subjects_to_keep) %>% droplevels() # then filter the fitted data points (data called in the second layer) my_plot$layers[[2]]$data <- my_plot$layers[[2]]$data %>% filter(subject %in% subjects_to_keep) %>% droplevels() # and finally, if 'displayCPS' was set to TRUE, filter the data used to display the CPS my_plot$layers[[4]]$data <- my_plot$layers[[4]]$data %>% filter(subject %in% subjects_to_keep) %>% droplevels() # plot the restricted my.plot my_plot #------ # It is also possible to export the curves in a pdf file running over several pages # and select the desired number of curves per page # set the desired number of subjects by page facet_nb = 4 # count the resulting number of pages num_pages = ceiling(length(unique(data_low_vision$subject))/facet_nb) # identify the list of subject names subjects_to_plot <- unique(as.character(data_low_vision$subject)) # split the list into chunks the same size as the number of subjects per page subjects_to_plot_splitted <- split(subjects_to_plot, ceiling(seq_along(subjects_to_plot)/facet_nb)) # create a pdf and wrap plots over several pages pdf("nlme-MNREAD-curves.pdf", width = 10.5, height = 8, paper="USr", useDingbats=T) for (i in seq(num_pages)) { my.plot <- nlmeCurve(nlme_model, displayCPS = F) # filter the original data points for the selected chunk of subjects my.plot$data <- my.plot$data %>% filter(subject %in% subjects_to_plot_splitted[[i]]) %>% droplevels() # filter the fitted data points for the selected chunk of subjects my.plot$layers[[2]]$data <- my.plot$layers[[2]]$data %>% filter(subject %in% subjects_to_plot_splitted[[i]]) %>% droplevels() print (my.plot + geom_line(colour = "red")) } dev.off() ``` ## See Also `nlmeModel` to fit MNREAD data using a nonlinear mixed-effect (NLME) modeling `nlmeParam` to estimate Maximum Reading Speed (MRS) and Critical Print Size (CPS) from the NLME model `mnreadCurve` for standard MNREAD curve

nlmeCurve function