added 'ylabel' argument to forecast_viz function, to display species …

…names on plot instead of sp codes

forecast_tools.R

@@ -324,17 +324,26 @@ all_forecasts$date=as.Date(all_forecasts$date)
 #' Visualize a time-series forecast
 #' Plots the observed time-series and the 1-step forecasts within it
 #' Plots the forecast time-series along with the prediction interval for future observations
-#' obs_data is a data.frame
-#' date_col_name is a string with the name for the date column
-#' val_col_name is a string with the name for the column of the value being forecast
+#' @param obs_data is a data.frame (observed data)
+#' @param obs_date_col_name is a string: name of the date column from obs_data
+#' @param obs_val_col_name is a string: name of the column of the value being forecast
+#' @param for_data is a data.frame (forecast data)
+#' @param for_date_col_name is a string: name of the date column from for_data
+#' @param for_val_col_name is a string: name of the column of value being forecast, from for_data
+#' @param for_model_name is a string: name of the model to be used from model column in for_data
+#' @param for_lowerpi_col_name is a string: name of the column of the lower confidence interval from for_data
+#' @param for_upperpi_col_name is a string: name of the column of the upper confidence interval from for_data
+#' @param start_newmoon is numeric: first new moon number to be plotted
+#' @param ylabel is a string: title for y-axis
 forecast_viz <- function(obs_data, obs_date_col_name, obs_val_col_name, for_data,
                          for_date_col_name, for_val_col_name, for_model_name,
-                         for_lowerpi_col_name, for_upperpi_col_name, start_newmoon){
+                         for_lowerpi_col_name, for_upperpi_col_name, start_newmoon,
+                         ylabel){
   for_data_sub = filter(for_data, species == obs_val_col_name, model == for_model_name)
   obs_data_sub = filter(obs_data, newmoonnumber >= start_newmoon)
   ggplot(obs_data_sub, aes_string(x = obs_date_col_name)) +
     geom_ribbon(data = for_data_sub, mapping = aes_string(x = for_date_col_name, ymin = for_lowerpi_col_name, ymax = for_upperpi_col_name), fill = "lightblue") +
     geom_line(aes_string(y = obs_val_col_name)) +
     geom_line(data = for_data_sub, mapping = aes_string(x = for_date_col_name, y = for_val_col_name), color = "blue") +
-    theme(axis.title.x=element_blank())
+    labs(x='',y=ylabel)
 }

index.Rmd

@@ -40,7 +40,8 @@ forecast_viz(obs_data = obs_data_newmoon,
              for_model_name = "ensemble",
              for_lowerpi_col_name = "LowerPI",
              for_upperpi_col_name = "UpperPI",
-             start_newmoon = 300)
+             start_newmoon = 300,
+             ylabel = 'Total Abundance')
 ```
 
 ## Species-Level Forecasts
@@ -61,17 +62,24 @@ most_abund_sp = sp_predictions %>%
   head(3) %>% 
   select(species)
 
-for (species in as.vector(most_abund_sp$species)) {
+# load in rodent species table to get scientific names to display on plots
+species_table = read.csv(
+  text = RCurl::getURL(
+      "https://raw.githubusercontent.com/weecology/PortalData/master/Rodents/Portal_rodent_species.csv"),stringsAsFactors = F,na.strings = '')
+most_abund_sp_names = filter(species_table,speciescode %in% most_abund_sp$species) %>% select(speciescode,scientificname)
+
+for (n in seq(dim(most_abund_sp_names)[1])) {
   species_forecast = forecast_viz(obs_data = obs_data_newmoon,
                obs_date_col_name = "censusdate",
-               obs_val_col_name = species,
+               obs_val_col_name = most_abund_sp_names$speciescode[n],
                for_data = for_data,
                for_date_col_name = "forecastdate",
                for_val_col_name = "estimate",
                for_model_name = "ensemble",
                for_lowerpi_col_name = "LowerPI",
                for_upperpi_col_name = "UpperPI",
-               start_newmoon = 300)
+               start_newmoon = 300,
+               ylabel = most_abund_sp_names$scientificname[n])
   plot(species_forecast)
 }
 ```