plotting_05_mice_ggplot.R

In earlycapistran/miceNls: Conservation Letters Utilities

# ---
# title: Mice ggplot
# author: Michelle María Early Capistrán 
# email:  earlycapistran@comunidad.unam.mx
# date: March 2021
# Script and data info:
#   - This script plots observed and multiply imputed values and trend lines
#     using results from ~/R/cf_mice_analysis.R and ~/R/cons_mice_analysis.R
#   - Trend lines and confidence intervals are plotted with results from 
#     ~/R/mice_pred_ci.R
#   - Data consists of green turtle (Chelonia mydas) CPUE data from monitoring
#     and derived from LEK.
#   - Data were obtained from ecological monitoring (1995-2018) in 
#     Bahía de los Ángeles, Baja California, Mexico by Dr. J. Seminoff, Grupo 
#     Tortuguero de Bahía de los Ángeles, and Comisión Nacional de Áreas 
#     Naturales Protegidas and LEK-derived data is from Early-Capistrán et al. 
#     (PeerJ, 2020)
#   - Multiply imputed data is generated by "~R/cons_mice_analysis.R" and 
#     "~R/cf_mice_analysis.R" and processed with "~/R/imputed_data_wrangling.R"
# - - -

# Install and load libraries --------------------------------------------------

# Check if required libraries are installed and install
# if necessary

# packages <- c("ggthemes", "patchwork", "tidyverse", "here")
# 
# if (length(setdiff(packages, rownames(installed.packages()))) > 0) {
#   install.packages(setdiff(packages, rownames(installed.packages())))
# }
# 
# devtools::install('consLettersUtils')

# Load libraries and packages
library("ggthemes")
library("tidyverse")
library("here")
library("devtools")
load_all("consLettersUtils")

# Source ggplot theme----------------------------------------------------------
# source("R/functions/theme_cmydas.R")

#  Load data and prepare data -------------------------------------------------
# Load observed and imputed data points from "~/R/imputed_data_wrangling.R"
obs_imp_data <- readRDS("results/obs_imp_data")

# Load pooled predicted values and confidence bounds
# Load observed and imputed data points generated with 
# "~/R/imputed_data_wrangling.R"
obs_imp_data <- readRDS("results/obs_imp_data")

# Load pooled predicted values and confidence bounds generated with 
# "~/R/mice_pred_ci.R"
cf_pred_ci <- readRDS("results/cf_pred_ci")
cons_pred_ci <- readRDS("results/cons_pred_ci")

# Start plotting -------------------------------------------------------------
# Store "year" values as label to replace serialized values
year_labels <- seq(1950, 2020, by=5)

# Plot pooled predicted values and confidence bounds
mice_ci <- ggplot()+
  geom_line(data = cons_pred_ci, aes(x = yearSerial, y = means), 
            color = "darkgrey") +
  geom_line(data = cons_pred_ci, aes(x = yearSerial, y = lwr), 
            linetype = 2, color = "darkgrey") +
  geom_line(data = cons_pred_ci, aes(x = yearSerial, y = upr), 
            linetype = 2, color = "darkgrey") +
  geom_line(data = cf_pred_ci, aes(x = yearSerial, y = means), 
            color = "darkgrey") +
  geom_line(data = cf_pred_ci, aes(x = yearSerial, y = lwr), 
            linetype = 2, color = "darkgrey") +
  geom_line(data = cf_pred_ci, aes(x = yearSerial, y = upr), 
            linetype = 2, color = "darkgrey")
mice_ci

 # Add scatter plot of observed values and mean imputed values
impMeanScatter <- mice_ci +
  geom_point(data= obs_imp_data %>% 
                     filter(type == "observed" | type == "imputed_means"),
             aes(x = yearSerial, y = cpue, shape = type, color = dataSource)) + 
  scale_shape_manual(labels = c("Mean imputed values", "Observed values"), 
                     values = c(1, 16)) +
  labs(y = "CPUE (turtles/12 hr)", 
       x = "Year", 
       color = "Source", 
       shape = "Data type",
       text=element_text(size=5)) +
  scale_color_tableau() +
  theme_cmydas()
impMeanScatter

# Add marginal rug plots to show density distribution of imputed values
miceRug <- impMeanScatter +
  geom_rug(data=obs_imp_data %>%   # Rug plot for imputed monitoring data
             filter(type == "imputed" & dataSource == "Monitoring"),
           aes(yearSerial, cpue),
           alpha = 0.05,
           position = "jitter", 
           color = "#F28E2B",
           sides ="r", 
           outside = FALSE) +
  geom_rug(data=obs_imp_data %>%   # Rug plot imputed for LEK data
           filter(type == "imputed" & dataSource == "LEK"),
         aes(yearSerial, cpue),
         alpha = 0.05,
         position = "jitter", 
         color = "#4E79A7",
         sides ="l",  
         outside = FALSE) +
  coord_cartesian(clip = "off") +
  theme(plot.margin = margin(0.5, 1.0, 0.5, 0.5, "cm")) 
miceRug

# Change x-axis labels to calendar years
miceRug <- miceRug +
  scale_x_continuous(
    breaks = seq(-2, 69, by=5),  # 1950 and 1985 in 'year'
    labels = year_labels,
    limits=c(-2, 69) # Adjust to display labels from 1950-1985
  )  
miceRug

# Save plot as svg
ggsave("figures/trend_plot.svg", device = svg)