# Run a `simmr_input` object through the Fixed Form Variational Bayes(FFVB) function This is the main function of simmr. It takes a `simmr_input` object created via `simmr_load`, runs it in fixed form Variational Bayes to determine the dietary proportions, and then outputs a `simmr_output` object for further analysis and plotting via `summary.simmr_output` and `plot.simmr_output`. ```r simmr_ffvb( simmr_in, prior_control = list(mu_0 = rep(0, simmr_in$n_sources), sigma_0 = 1), ffvb_control = list(n_output = 3600, S = 100, P = 10, beta_1 = 0.9, beta_2 = 0.9, tau = 100, eps_0 = 0.0225, t_W = 50) ) ``` ## Arguments - `simmr_in`: An object created via the function `simmr_load` - `prior_control`: A list of values including arguments named `mu_0` (prior for mu), and `sigma_0` (prior for sigma). - `ffvb_control`: A list of values including arguments named `n_output` (number of rows in theta output), `S` (number of samples taken at each iteration of the algorithm), `P` (patience parameter), `beta_1` and `beta_2` (adaptive learning weights), `tau` (threshold for exploring learning space), `eps_0` (fixed learning rate), `t_W` (rolling window size) ## Returns An object of class `simmr_output` with two named top-level components: - **input**: The `simmr_input` object given to the `simmr_ffvb` function - **output**: A set of outputs produced by the FFVB function. These can be analysed using the `summary.simmr_output` and `plot.simmr_output` functions. ## Examples ```r ## Not run: ## See the package vignette for a detailed run through of these 4 examples # Data set 1: 10 obs on 2 isos, 4 sources, with tefs and concdep data(geese_data_day1) simmr_1 <- with( geese_data_day1, simmr_load( mixtures = mixtures, source_names = source_names, source_means = source_means, source_sds = source_sds, correction_means = correction_means, correction_sds = correction_sds, concentration_means = concentration_means ) ) # Plot plot(simmr_1) # Print simmr_1 # FFVB run simmr_1_out <- simmr_ffvb(simmr_1) # Print it print(simmr_1_out) # Summary summary(simmr_1_out, type = "correlations") summary(simmr_1_out, type = "statistics") ans <- summary(simmr_1_out, type = c("quantiles", "statistics")) # Plot plot(simmr_1_out, type = "boxplot") plot(simmr_1_out, type = "histogram") plot(simmr_1_out, type = "density") plot(simmr_1_out, type = "matrix") # Compare two sources compare_sources(simmr_1_out, source_names = c("Zostera", "Enteromorpha")) # Compare multiple sources compare_sources(simmr_1_out) ##################################################################################### # A version with just one observation data(geese_data_day1) simmr_2 <- with( geese_data_day1, simmr_load( mixtures = mixtures[1, , drop = FALSE], source_names = source_names, source_means = source_means, source_sds = source_sds, correction_means = correction_means, correction_sds = correction_sds, concentration_means = concentration_means ) ) # Plot plot(simmr_2) # FFVB run - automatically detects the single observation simmr_2_out <- simmr_ffvb(simmr_2) # Print it print(simmr_2_out) # Summary summary(simmr_2_out) ans <- summary(simmr_2_out, type = c("quantiles")) # Plot plot(simmr_2_out) plot(simmr_2_out, type = "boxplot") plot(simmr_2_out, type = "histogram") plot(simmr_2_out, type = "density") plot(simmr_2_out, type = "matrix") ##################################################################################### # Data set 2: 3 isotopes (d13C, d15N and d34S), 30 observations, 4 sources data(simmr_data_2) simmr_3 <- with( simmr_data_2, simmr_load( mixtures = mixtures, source_names = source_names, source_means = source_means, source_sds = source_sds, correction_means = correction_means, correction_sds = correction_sds, concentration_means = concentration_means ) ) # Get summary print(simmr_3) # Plot 3 times plot(simmr_3) plot(simmr_3, tracers = c(2, 3)) plot(simmr_3, tracers = c(1, 3)) # See vignette('simmr') for fancier axis labels # FFVB run simmr_3_out <- simmr_ffvb(simmr_3) # Print it print(simmr_3_out) # Summary summary(simmr_3_out) summary(simmr_3_out, type = "quantiles") summary(simmr_3_out, type = "correlations") # Plot plot(simmr_3_out) plot(simmr_3_out, type = "boxplot") plot(simmr_3_out, type = "histogram") plot(simmr_3_out, type = "density") plot(simmr_3_out, type = "matrix") ################################################################ # Data set 5 - Multiple groups Geese data from Inger et al 2006 # Do this in raw data format - Note that there's quite a few mixtures! data(geese_data) simmr_5 <- with( geese_data, simmr_load( mixtures = mixtures, source_names = source_names, source_means = source_means, source_sds = source_sds, correction_means = correction_means, correction_sds = correction_sds, concentration_means = concentration_means, group = groups ) ) # Plot plot(simmr_5, xlab = expression(paste(delta^13, "C (per mille)", sep = "")), ylab = expression(paste(delta^15, "N (per mille)", sep = "")), title = "Isospace plot of Inger et al Geese data" ) # Run MCMC for each group simmr_5_out <- simmr_ffvb(simmr_5) # Summarise output summary(simmr_5_out, type = "quantiles", group = 1) summary(simmr_5_out, type = "quantiles", group = c(1, 3)) summary(simmr_5_out, type = c("quantiles", "statistics"), group = c(1, 3)) # Plot - only a single group allowed plot(simmr_5_out, type = "boxplot", group = 2, title = "simmr output group 2") plot(simmr_5_out, type = c("density", "matrix"), grp = 6, title = "simmr output group 6") # Compare sources within a group compare_sources(simmr_5_out, source_names = c("Zostera", "U.lactuca"), group = 2) compare_sources(simmr_5_out, group = 2) # Compare between groups compare_groups(simmr_5_out, source = "Zostera", groups = 1:2) compare_groups(simmr_5_out, source = "Zostera", groups = 1:3) compare_groups(simmr_5_out, source = "U.lactuca", groups = c(4:5, 7, 2)) ## End(Not run) ``` ## References Andrew C. Parnell, Donald L. Phillips, Stuart Bearhop, Brice X. Semmens, Eric J. Ward, Jonathan W. Moore, Andrew L. Jackson, Jonathan Grey, David J. Kelly, and Richard Inger. Bayesian stable isotope mixing models. Environmetrics, 24(6):387–399, 2013. Andrew C Parnell, Richard Inger, Stuart Bearhop, and Andrew L Jackson. Source partitioning using stable isotopes: coping with too much variation. PLoS ONE, 5(3):5, 2010. ## See Also `simmr_load` for creating objects suitable for this function, `plot.simmr_input` for creating isospace plots, `summary.simmr_output` for summarising output, and `plot.simmr_output` for plotting output. ## Author(s) Andrew Parnell , Emma Govan

simmr_ffvb function