Sample script to create ggplots incorporating RMSE, MBE and NSE (based on COD method) plus an example of how to create an animated plot and interactive plot

Sample Plotting Example.R

# Libraries ----
library(dplyr)
library(ggplot2) # Plotting data
library(ggrepel) # Prevent overlapping text labels
library(Metrics) # Calculate Statistics
library(plotly) # Interactive plots
library(gganimate) # Animate plots

# Sample Statistical Plots ----
# Create sample data
set.seed(123) # For reproducibility
sample_data <- data.frame(
  source = rep(c("Source A", "Source B"), each = 10),
  observed_data = c(rnorm(10, mean = 8000, sd = 500), rnorm(10, mean = 8500, sd = 600)),
  model_data = c(rnorm(10, mean = 8200, sd = 500), rnorm(10, mean = 8600, sd = 600)),
  time = rep(1:10, 2)
)

# Simplified RMSE, NSE, and MBE calculations in one step
metrics_data <- sample_data %>%
  group_by(source) %>%
  summarise(
    rmse = Metrics::rmse(observed_data, model_data),
    nse = 1 - sum((model_data - observed_data)^2) / sum((observed_data - mean(observed_data))^2),
    mbe = mean(model_data - observed_data)
  )

# Determine the range for axes
range_vals <- range(c(sample_data$model_data, sample_data$observed_data))

# Plot the data
ggplot(sample_data, aes(x = model_data, y = observed_data, colour = source)) +
  geom_point(size = 4) +
  geom_smooth(method = "lm", se = FALSE) +
  geom_abline(intercept = 0, slope = 1, linetype = "dashed") +
  labs(x = "Modelled Values", y = "Observed Values", colour = "Source") +
  geom_text_repel(
    data = metrics_data,
    aes(x = range_vals[2], y = range_vals[2], label = paste("RMSE:", round(rmse, 2))),
    hjust = 1, vjust = 1, direction = "y"
  ) +
  geom_text_repel(
    data = metrics_data,
    aes(x = range_vals[1], y = range_vals[2], label = paste("NSE:", round(nse, 2))),
    hjust = 0, vjust = 1, direction = "y"
  ) +
  geom_text_repel(
    data = metrics_data,
    aes(x = range_vals[1], y = range_vals[1], label = paste("MBE:", round(mbe, 2))),
    hjust = 0, vjust = 0, direction = "y"
  ) +
  coord_equal() +
  theme_bw() +
  lims(x = range_vals, y = range_vals) +
  # ggtitle("Sample Data") +
  guides(
    colour = guide_legend(override.aes = list(linetype = NA, size = 4)),
    shape = guide_legend(override.aes = list(size = 4))
  )

# Animated Plotting ----
# Create the ggplot object
p <- ggplot(sample_data, aes(x = observed_data, y = model_data, color = source)) +
  geom_point(size = 3) +
  labs(
    title = "Observed vs. Model Data",
    subtitle = "Time: {frame_time}",
    x = "Observed Data",
    y = "Model Data"
  ) +
  theme_minimal() +
  transition_time(time) +
  ease_aes("linear")

# Animate the plot
animate(p, nframes = 100, fps = 10, width = 700, height = 700)

# Save the animation as a GIF (optional)
anim_save("observed_vs_model_animation.gif", animation = last_animation())

# Interactive Plotting ----
q <- ggplot(sample_data, aes(x = model_data, y = observed_data, colour = source)) +
  geom_point(size = 4) +
  geom_smooth(method = "lm", se = FALSE) +
  geom_abline(intercept = 0, slope = 1, linetype = "dashed") +
  labs(x = "Modelled Values", y = "Observed Values", colour = "Source") +
  coord_equal() +
  theme_bw() +
  lims(x = range_vals, y = range_vals) +
  ggtitle("Sample Data") +
  guides(
    colour = guide_legend(override.aes = list(linetype = NA, size = 4)),
    shape = guide_legend(override.aes = list(size = 4))
  )

ggplotly(q)