In pratikunterwegs/patho-move-evol: Analysis code for the _pathomove_ model.

Plot strategy-specific differences in ecological outcomes

library(data.table)
library(ggplot2)
library(patchwork)
library(colorspace)

Load strategy-wise outcome data

df_strat <- fread("data/results/data_default_strategy_evo.csv")

df_strat <- df_strat[, c(
  "gen", "gen_abs", "social_strat",
  "mean_move", "mean_assoc", "mean_intake",
  "mean_energy"
)]

# cast long
df_strat <- melt(df_strat, id.vars = c("gen", "gen_abs", "social_strat"))

# remove non-handler tracking
df_strat <- df_strat[social_strat != "non-handler tracking"]

# log strategy data
sgen <- 3000
genmax <- 5000

Plot strategy differences in movement

# plot strategy differences in movement distance
plot_diff_move <-
  ggplot(df_strat[variable %in% c("mean_move") &
    gen %% 100 == 0]) +
  geom_vline(
    xintercept = c(0),
    lty = 2,
    size = 0.3,
    col = c("red")
  ) +
  stat_summary(
    aes(gen_abs, value, col = social_strat),
    geom = "line", size = 0.2, show.legend = FALSE
  ) +
  stat_summary(
    aes(gen_abs, value, col = social_strat),
    size = 0.2,
    shape = 1, show.legend = FALSE
  ) +
  scale_x_continuous(
    breaks = c(-500, 0, 500, 2000),
    labels = scales::comma_format(
      accuracy = 1
    ),
    name = "Gens. after pathogen intro.",
    sec.axis = sec_axis(
      trans = function(x) x + sgen,
      breaks = c(sgen, genmax),
      labels = scales::comma_format(
        accuracy = 1
      ),
      name = "Generations"
    )
  ) +
  scale_colour_discrete_sequential(
    palette = "Viridis",
    l2 = 80,
    rev = FALSE,
    name = NULL,
    limits = c("agent avoiding", "agent tracking", "handler tracking"),
    labels = stringr::str_to_sentence,
    na.value = "lightgrey"
  ) +
  coord_cartesian(
    xlim = c(-500, 2000)
  ) +
  theme_test(
    base_family = "Arial",
    base_size = 8
  ) +
  theme(
    legend.position = "bottom",
    legend.key = element_blank(),
    legend.background = element_blank(),
    strip.background = element_blank(),
    strip.text = element_text(
      face = "italic"
    ),
    plot.background = element_blank(),
    strip.placement = "outside"
  ) +
  labs(
    x = NULL,
    y = "Distance moved",
    colour = NULL,
    shape = NULL
  ) +
  guides(
    shape = NULL,
    colour = guide_legend(
      override.aes = list(
        shape = 15
      )
    )
  )

Plot strategy differences in intake

# plot strategy differences in intake
plot_diff_intake <-
  ggplot(df_strat[variable %in% c("mean_intake") &
    gen %% 100 == 0]) +
  geom_vline(
    xintercept = c(0),
    lty = 2,
    size = 0.3,
    col = c("red")
  ) +
  stat_summary(
    aes(gen_abs, value, col = social_strat),
    geom = "line", size = 0.2, show.legend = FALSE
  ) +
  stat_summary(
    aes(gen_abs, value, col = social_strat),
    size = 0.2,
    shape = 2, show.legend = FALSE
  ) +
  scale_x_continuous(
    breaks = c(-500, 0, 500, 2000),
    labels = scales::comma_format(
      accuracy = 1
    ),
    name = "Gens. after pathogen intro.",
    sec.axis = sec_axis(
      trans = function(x) x + sgen,
      breaks = c(sgen, genmax),
      labels = scales::comma_format(
        accuracy = 1
      ),
      name = "Generations"
    )
  ) +
  scale_colour_discrete_sequential(
    palette = "Viridis",
    l2 = 80,
    rev = FALSE,
    name = NULL,
    limits = c("agent avoiding", "agent tracking", "handler tracking"),
    labels = stringr::str_to_sentence,
    na.value = "lightgrey"
  ) +
  coord_cartesian(
    xlim = c(-500, 2000)
  ) +
  theme_test(
    base_family = "Arial",
    base_size = 8
  ) +
  theme(
    legend.position = "bottom",
    legend.key = element_blank(),
    legend.background = element_blank(),
    strip.background = element_blank(),
    strip.text = element_text(
      face = "italic"
    ),
    plot.background = element_blank(),
    strip.placement = "outside"
  ) +
  labs(
    x = NULL,
    y = "Per-capita intake",
    colour = NULL,
    shape = NULL
  ) +
  guides(
    shape = NULL,
    colour = guide_legend(
      override.aes = list(
        shape = 15
      )
    )
  )

Plot strategy differences in energy

# plot strategy differences in energy
plot_diff_energy <-
  ggplot(df_strat[variable %in% c("mean_energy") & gen %% 100 == 0]) +
  geom_vline(
    xintercept = c(0),
    lty = 2,
    size = 0.3,
    col = c("red")
  ) +
  stat_summary(
    aes(gen_abs, value, col = social_strat),
    geom = "line", size = 0.2, show.legend = FALSE
  ) +
  stat_summary(
    aes(gen_abs, value, col = social_strat),
    size = 0.2,
    shape = 5, show.legend = FALSE
  ) +
  scale_x_continuous(
    breaks = c(-500, 0, 500, 2000),
    labels = scales::comma_format(
      accuracy = 1
    ),
    name = "Gens. after pathogen intro.",
    sec.axis = sec_axis(
      trans = function(x) x + sgen,
      breaks = c(sgen, genmax),
      labels = scales::comma_format(
        accuracy = 1
      ),
      name = "Generations"
    )
  ) +
  scale_colour_discrete_sequential(
    palette = "Viridis",
    l2 = 80,
    rev = FALSE,
    name = NULL,
    limits = c("agent avoiding", "agent tracking", "handler tracking"),
    labels = stringr::str_to_sentence,
    na.value = "lightgrey"
  ) +
  coord_cartesian(
    xlim = c(-500, 2000),
    ylim = c(-6, NA)
  ) +
  theme_test(
    base_family = "Arial",
    base_size = 8
  ) +
  theme(
    legend.position = "bottom",
    legend.key = element_blank(),
    legend.background = element_blank(),
    strip.background = element_blank(),
    strip.text = element_text(
      face = "italic"
    ),
    plot.background = element_blank(),
    strip.placement = "outside"
  ) +
  labs(
    x = NULL,
    y = "Per-capita energy",
    colour = NULL,
    shape = NULL
  ) +
  guides(
    shape = NULL,
    colour = guide_legend(
      override.aes = list(
        shape = 15
      )
    )
  )

Prepare combined figures

plot_strat_diff <- wrap_plots(
  plot_diff_move, plot_diff_intake, plot_diff_energy,
  ncol = 3, guides = "collect"
) & theme(
  legend.position = "none"
)

Plot associations per distance moved by strategy

Load data on associations per movement

files <- list.files(
  path = "data/results/move_assoc_data",
  pattern = "default", full.names = TRUE
)

# read data and filter for default paramaeter combination
data_move_assoc <- lapply(files, fread) |> rbindlist()
data_move_assoc <- data_move_assoc[between(gen, 3000, 3500), ]
data_move_assoc <- data_move_assoc[costInfect == 0.25 &
  regen_time == 50, ]

# round moved
data_move_assoc[, moved_bin := plyr::round_any(moved, 5)]
data_move_assoc <- data_move_assoc[social_strat != "non-handler tracking", ]

# plot associations per movement distance
plot_move_assoc <-
  ggplot() +
  geom_jitter(
    data = data_move_assoc[gen %% 10 == 0, ],
    aes(moved, mean_assoc, col = social_strat),
    size = 0.1, alpha = 0.5
  ) +
  geom_vline(
    xintercept = 50,
    linewidth = 0.3, lty = "dashed", colour = "indianred"
  ) +
  stat_summary(
    data = data_move_assoc,
    aes(moved_bin, mean_assoc,
      fill = social_strat
    ),
    shape = 21
  ) +
  annotate(
    geom = "text",
    x = 100,
    y = 750,
    label = "Post-pathogen\n3000 ≤ G ≤ 3500",
    hjust = "inward",
    fontface = "italic",
    colour = "red",
    size = 3,
    family = "Arial"
  ) +
  scale_fill_discrete_sequential(
    palette = "Viridis",
    l1 = 15, l2 = 80,
    rev = F,
    breaks = c("agent avoiding", "handler tracking", "agent tracking"),
    name = NULL,
    labels = stringr::str_to_sentence
  ) +
  scale_colour_discrete_sequential(
    palette = "Viridis",
    c1 = 30, c2 = 40,
    l1 = 50, l2 = 80,
    rev = F,
    name = NULL,
    labels = stringr::str_to_sentence
  ) +
  scale_y_continuous(
    labels = scales::comma_format()
  ) +
  coord_cartesian(
    ylim = c(30, 1000),
    expand = T
  ) +
  theme_test(
    base_family = "Arial",
    base_size = 8
  ) +
  theme(
    axis.text.y = element_text(
      angle = 90,
      hjust = 0.5
    ),
    legend.key.height = unit(1, "mm"),
    legend.position = "bottom"
  ) +
  labs(
    x = "Distance moved",
    y = "Encounters w/ individuals"
  ) +
  guides(
    alpha = "none",
    shape = "none",
    colour = "none",
    fill = guide_legend(
      override.aes = list(
        alpha = 1
      )
    )
  )

Plot infection rate per strategy

Load data on infection rate per strategy

# load data
data <- fread("data/results/data_default_strategy_evo.csv")
data <- data[between(gen, 3000, 3500), ]

# assign broad strategy
data[, social_strat := fifelse(
  social_strat %in% c("agent avoiding", "handler tracking"),
  social_strat, "other"
)]

# assign levels
data[, social_strat := factor(social_strat,
  levels = c(
    "other", "handler tracking",
    "agent avoiding"
  )
)]

# plot infection rate histogram per strategy
plot_infected <-
  ggplot(data) +
  ggdist::stat_histinterval(
    aes(
      social_strat, prop_infec,
      fill = social_strat,
      group = social_strat
    ),
    size = 0.5,
    n = 31,
    show.legend = F
  ) +
  scale_fill_discrete_sequential(
    palette = "Viridis",
    l2 = 80,
    order = c(2, 1, 3),
    name = NULL
  ) +
  scale_y_continuous(
    labels = scales::percent
  ) +
  scale_x_discrete(
    breaks = c(
      "agent avoiding",
      "handler tracking",
      "other"
    ),
    limits = c(
      "agent avoiding",
      "handler tracking",
      "other"
    )
  ) +
  theme_test(
    base_family = "Arial",
    base_size = 8
  ) +
  theme(
    legend.position = "top",
    legend.key.height = unit(0.5, "mm"),
    legend.key.width = unit(2, "mm"),
    axis.text.y = element_text(
      angle = 90,
      hjust = 0.5
    ),
    plot.background = element_blank(),
    axis.text.x = element_blank()
  ) +
  labs(
    x = "Social movement\nstrategy",
    y = "% Infected"
  )

Prepare figure for strategy-wise ecological differences

# prepare and plot figure for strategy-wise social differences
plot_soc_out <-
  wrap_plots(
    plot_strat_diff,
    plot_move_assoc,
    plot_infected,
    guides = "collect",
    design = "AAA\nBBC\nBBC"
  ) &
    plot_annotation(
      tag_levels = c("A", 1)
    ) &
    theme(
      legend.position = "bottom",
      plot.tag = element_text(
        face = "bold"
      )
    )

ggsave(
  plot = plot_soc_out,
  file = "figures/fig_social_outcomes.png",
  height = 120, width = 120,
  units = "mm"
)