synchronize class colors b/n map & time series plot

[bbest]

From Jenn Brown:

Hi again all,
I was back looking at the Seascapes for Sanctuaries Shiny app. I was looking at the outputs for Florida Keys NMS and was confused for a bit by the color scheme in the map vs. the time series. See this screenshot as an example.

My assumption that the color for a given seascape class was the SAME in the map and in the time series. However, as I tried to figure out why the peachy-orange color is prominent on the map but missing from the time series, I realized that the seascape class color schemes are DIFFERENT - seascape class 15 is peachy-orange on the map and purple on the time series.

I think it would be easier for a user to make sense of how the mapping output relates to the time series output if each seascape class has the same color on the map and in the time series.

I don't know if this is an easy or difficult change, but wanted to bring it to your attention.

[bbest]

It shouldn't be too hard to do. Here are the steps needed to bake a consistent color palette into the mapping and time series plotting functions in seascapeR also used by the Seascapes for Sanctuaries Shiny app (principally in server.R):

1. Colors. Define common color ramp, e.g. based on "Spectral", coded as new column color to each class in the ss_gl_classes dataset.
2. Map.
   1. Update map_ss_grd() with ability to apply a custom categorical color scheme instead of simply grd_palette argument.
   2. Fix map portion of app here to use scheme with the leaflet proxy object:
      

      seascapes-app/app/server.R

      Lines 115 to 121 in 27059cc

      	addRasterImage(
      	grd,
      	project = T, method = "ngb",
      	colors = pal,
      	opacity = 0.8,
      	layerId = "grd",
      	group = "Class") %>%

3. Time Series Plot
   1. Update plot_ss_ts() with arguments for optionally applying custom color scheme (or palette).
   2. Update time series plot of app here:
      

      seascapes-app/app/server.R

      Lines 176 to 178 in 27059cc

      	output$plot <- renderDygraph({
      	g <- plot_ss_ts(values$tbl, values$col)

I should be able to attend to this by the end of the month.