Visualization

[kaydin]

Share pretty plots and ideas for plots!!

[biadias]

Function for making the rsim.plot() interactive. The base function is pretty much a copy of the rsim.plot() and then I called the plotly object. It is highly customized, and I would welcome anybody to make it more general.
Maybe it is not very useful inside of the package, because it depends on tidyverse and plotly.

`# --- # --- # --- # --- # --- # --- # --- # --- # --- # 
`# Interactive plots for Rsim plot using plotly:: package and tidyverse::
`# Author: Bia Dias
`# Date: 12/13/2024
`# --- # --- # --- # --- # --- # --- # --- # --- # --- # 

library(plotly)
library(tidyverse)
library(colorspace) # default color palette, but I can simplify this later
library(grDevices)


rsim.plot.interactive <- function(Rsim.output, spname="all", indplot = FALSE, palette= "rainbow") {
  rsim_name <- deparse(substitute(Rsim.output))
  mrg <- list(l = 50, r = 50,
              b = 50, t = 50,
              pad = 20)
  
  if ("all" %in% spname) {
    spname <- colnames(Rsim.output$out_Biomass)[2:ncol(Rsim.output$out_Biomass)]
  }
  # Extract and calculate rel. B
  if (!indplot) {
    # Multiple functional groups
    biomass <- Rsim.output$out_Biomass[, spname, drop = FALSE]
    start.bio <- biomass[1,]
    start.bio[start.bio == 0] <- 1
    rel.bio <- sweep(biomass, 2, start.bio, "/") # divide each column by its start value
  } else {
    # For a single functional group
    spnum <- which(Rsim.output$params$spname == spname)
    biomass <- Rsim.output$out_Biomass[, spnum]
    rel.bio <- biomass / biomass[1]
    rel.bio <- matrix(rel.bio, ncol = 1, dimnames = list(NULL, spname))
  }
  
  # Create a time vector
  time <- seq_len(nrow(rel.bio))
  df <- cbind(time, as.data.frame(rel.bio))
  df_long <- tidyr::pivot_longer(df, cols = -time, names_to = "Species", values_to = "RelativeBiomass")
  df_long$Species <- factor(df_long$Species, levels = spname)
  n_species <- length(spname)
  #b_palette <- colorspace::rainbow_hcl 
  
  if (is.character(palette) && length(palette) == 1 && exists(palette, mode = "function")) {
    # If the palette is pre-determined e.g. "rainbow" or "terrain.colors" 
    # https://www.nceas.ucsb.edu/sites/default/files/2020-04/colorPaletteCheatsheet.pdf 
    pal_fun <- match.fun(palette)
    my_colors <- pal_fun(n_species)
  } else if (is.character(palette) && length(palette) > 1) {
    # If you decide to give a vector of colors
    if (length(palette) < n_species) {
      # the function will interpolate if there not enough colors
      my_colors <- grDevices::colorRampPalette(palette)(n_species)
    } else {
      # If enough colors or more
      my_colors <- palette[1:n_species]
    }
  } else {
    my_colors <- rainbow(n_species)
  }
  
  # Plotly object  
  rsim.int.plotly <- plot_ly(data = df_long,
               x = ~time,
               y = ~RelativeBiomass,
               color = ~Species,
               colors = my_colors,
               type = 'scatter',
               mode = 'lines') %>%
    layout(title = paste("Relative Biomass Over Time:" , rsim_name),
           xaxis = list(title = "Months"),
           yaxis = list(title = "Relative Biomass"),
           margin=mrg)
  
  return(rsim.int.plotly)
}

[kaydin]

Moving to package rpathviz