Activity 18 — Mapping Redux: Ranges That Move

Time on a map with facets, convex hulls and hex grids, two species at once, and an interactive ggplotly map

maps

spatial

gbif

ggplotly

plotly

invasive species

Hands-on companion to the mapping redux lecture. Students download spotted lanternfly records one snapshot year at a time, facet them to put time on the map, turn the dots into a range with a convex hull and then a hex grid, watch the hull fail on the Eurasian collared-dove, map two flying squirrel species at once, and finish with an interactive ggplotly map with a period slider. Homework is the same recipe on a species they pick.

Author

Bill Perry

Published

November 10, 2026

In-class Activity 18: Ranges That Move 🪰🕊️🐿️

Today’s Objectives

Download occurrence records one snapshot year at a time, and know why
Put time on a map with facet_wrap()
Turn a cloud of dots into a range with a convex hull — and catch it lying
Build a hex grid, join points into it, and count
Map two species at once
Make a map interactive with ggplotly(), including a period slider
Do it again on a species you pick, for homework

⚠️ Install this BEFORE class

One new package:

install.packages("plotly")

📂 Files for today

In-class skeleton: 18_mapping_redux_skeleton.R → save it as scripts/18_mapping_redux.R. All of today’s code is already in it.
Homework skeleton: 18_mapping_redux_homework.R → save it as scripts/18_mapping_redux_homework.R. This one is what you turn in.
Data — all of these go in your data/ folder:
- collared_dove_gbif.csv
- flying_squirrels_gbif.csv
- opossum_gbif.csv
- giant_swallowtail_gbif.csv
- spotted_lanternfly_gbif.csv — a backup, in case the live download fails

What you hand in

Today is a run-it-and-look-at-it day, like last mapping class. Every line is in this activity and in the skeleton. You do not turn in scripts/18_mapping_redux.R.
The Q boxes are things to think about. Not graded, not written down.
You turn in scripts/18_mapping_redux_homework.R — a species you pick, mapped across time.

🧩 Chunk 1 — Time on a map (after lecture Chunk 1)

Parts 1–4: download year by year, stack, and facet.

Part 1 · Load up and download three snapshot years

▶ Run this:

# Load libraries ---------------------------------------
library(tidyverse)
library(sf)       # spatial data frames
library(maps)     # ready-made country and state outlines
library(rgbif)    # download species records from GBIF
library(plotly)   # interactive plots

Last time we asked GBIF for 2,000 moose records in one go and they all came back from a single year. That was a curiosity then. Today it would wreck everything — so we ask for one year at a time.

▶ Run this (it takes about 20 seconds):

# One download per snapshot year -----------------------
slf_2018 <- occ_data(scientificName = "Lycorma delicatula", country = "US",
                     hasCoordinate = TRUE, year = 2018, limit = 1500)$data

slf_2021 <- occ_data(scientificName = "Lycorma delicatula", country = "US",
                     hasCoordinate = TRUE, year = 2021, limit = 1500)$data

slf_2025 <- occ_data(scientificName = "Lycorma delicatula", country = "US",
                     hasCoordinate = TRUE, year = 2025, limit = 1500)$data

Three identical calls. Only year changes.

⚠️ Watch out! If a download fails — bad wifi, GBIF having a bad day — skip ahead to Part 14. There is a cached copy of all of this in your data/ folder and the rest of the activity will work fine from it.

✏️ Q1: The same limit = 1500 is on all three calls. Is that a target or a ceiling? What do you expect 2018 to come back with, given the insect was first found in 2014?

Part 2 · Stack them and look hard at the counts

▶ Run this:

# Stack the three years, keep five columns -------------
slf_df <- bind_rows(slf_2018, slf_2021, slf_2025) %>%
  select(species,
         state     = stateProvince,
         year,
         longitude = decimalLongitude,
         latitude  = decimalLatitude) %>%
  filter(!is.na(longitude), !is.na(latitude))

count(slf_df, year)

bind_rows() stacks data frames that share column names — the three years become one table
select() renames while it selects: state = stateProvince means new name = old name
GBIF’s coordinate columns get short names here so the rest of the day is easier to type

✏️ Q2: One of those three years hit the ceiling of 1,500 and one did not. Which is which, and what does that tell you about how many records exist?

✏️ Q3: GBIF holds roughly 21,000 US lanternfly records for 2025. You have 1,500 of them. If more dots show up in the 2025 panel than the 2018 panel, does that prove there were more lanternflies?

💡 Key idea: you cannot compare the number of dots across panels — the download caps saw to that. You can compare where they are. Shape, not count.

Part 3 · Make it spatial

Nothing new here. This is the exact recipe from the Bigfoot map.

▶ Run this:

# lon/lat -> geometry ----------------------------------
slf_sf <- slf_df %>%
  st_as_sf(coords = c("longitude", "latitude"), crs = 4326)

# Basemap: the lower 48 --------------------------------
states_sf <- st_as_sf(
  maps::map("state", plot = FALSE, fill = TRUE)
) %>%
  st_set_crs(4326)

slf_sf

✏️ Q4: slf_df had a longitude and a latitude column. Look at slf_sf — where did they go?

Part 4 · Put time on the map

🔮 Predict first: three panels, 2018 / 2021 / 2025. Sketch on paper what you think each one looks like before you run it.

▶ Run this:

# One panel per snapshot year --------------------------
ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = slf_sf, color = "firebrick", size = 0.7, alpha = 0.5) +
  coord_sf(xlim = c(-92, -69), ylim = c(33, 45)) +
  facet_wrap(~ year) +
  theme_void()

facet_wrap(~ year) is the only new line. It is the same faceting you used on penguins and on leaves — a map is just a plot.

✏️ Q5: Find the 2018 cluster. That is Berks County, Pennsylvania, where the insect was first found in 2014. Which direction did it go first?

✏️ Q6: Pull up a map of interstate highways in your head, or in a browser. Does the 2025 spread look like something that walked, or something that hitched a ride?

💡 Key idea: small multiples. Everything is held constant — same basemap, same limits, same colours — except the one thing you are comparing. Side by side beats an animation, because you can look instead of having to remember.

🧩 Chunk 2 — From dots to a range (after lecture Chunk 2)

Parts 5–7: the convex hull, and the day it lies.

Part 5 · The convex hull

A convex hull is the smallest polygon that contains every point — stretch a rubber band around the whole cloud and let go.

▶ Run this:

# One hull per year ------------------------------------
slf_hulls <- slf_sf %>%
  group_by(year) %>%
  summarise(geometry = st_union(geometry)) %>%
  st_convex_hull()

slf_hulls

group_by() then summarise() — the same two verbs you have used since week three
st_union() glues one year’s points into a single multipoint blob
st_convex_hull() wraps the rubber band around that blob

▶ Run this:

# Hull behind, points on top ---------------------------
ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = slf_hulls, aes(fill = factor(year)), alpha = 0.35, color = "grey30") +
  geom_sf(data = slf_sf, size = 0.4, alpha = 0.3) +
  coord_sf(xlim = c(-92, -69), ylim = c(33, 45)) +
  facet_wrap(~ year) +
  theme_void() +
  theme(legend.position = "none")

▶ Run this:

# Put a number on it -----------------------------------
slf_hulls %>%
  mutate(area_km2 = as.numeric(st_area(geometry)) / 1e6) %>%
  st_drop_geometry()

st_area() measures a polygon. It reports square metres, so divide by a million for km²
st_drop_geometry() throws the geometry column away and leaves a plain table

✏️ Q7: Look at the 2025 hull. Put your finger on a big patch of it that has no dots in it at all. Is the lanternfly there?

✏️ Q8: The area roughly triples between panels. Is “the range tripled” a fair sentence to write about those numbers?

Part 6 · A second species — the Eurasian collared-dove

The Eurasian collared-dove, Streptopelia decaocto, escaped a pet shop in the Bahamas in the 1970s, reached Florida in the early 1980s, and crossed the whole continent in about thirty years.

This file was downloaded for you — one capped batch per year from 1990 to 2025, which takes 36 separate downloads and far longer than a class period. The script that did it is in scripts/00_build_cached_data.R if you want to see it.

▶ Run this:

# Read the cached records ------------------------------
dove_df <- read_csv("data/collared_dove_gbif.csv")

# Cut 36 years into four periods -----------------------
dove_df <- dove_df %>%
  mutate(period = case_when(
    year <= 1998 ~ "1990-1998",
    year <= 2006 ~ "1999-2006",
    year <= 2014 ~ "2007-2014",
    TRUE         ~ "2015-2025"))

count(dove_df, period)

case_when() — straight out of the factors module, doing exactly what it did there
Periods instead of single years, because one year of 1990s bird records is far too thin
The labels are written so they sort alphabetically into chronological order. That is deliberate — it is why the panels will come out in the right order with no extra work

▶ Run this:

# Spatial, then the dots -------------------------------
dove_sf <- dove_df %>%
  st_as_sf(coords = c("longitude", "latitude"), crs = 4326)

ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = dove_sf, color = "firebrick", size = 0.4, alpha = 0.4) +
  coord_sf(xlim = c(-125, -66), ylim = c(24, 50)) +
  facet_wrap(~ period, nrow = 1) +
  theme_void()

✏️ Q9: Describe the four panels in one sentence each. Where does the bird start, and where does it end up?

Part 7 · Now watch the hull fail

Same two lines that worked on the lanternfly. Same species of question. Completely different result.

▶ Run this:

# The identical hull recipe ----------------------------
dove_hulls <- dove_sf %>%
  group_by(period) %>%
  summarise(geometry = st_union(geometry)) %>%
  st_convex_hull()

ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = dove_hulls, fill = "steelblue", alpha = 0.3, color = "steelblue4") +
  geom_sf(data = dove_sf, color = "firebrick", size = 0.3, alpha = 0.3) +
  coord_sf(xlim = c(-125, -66), ylim = c(24, 50)) +
  facet_wrap(~ period, nrow = 1) +
  theme_void()

▶ Run this:

# How many records are west of 100 degrees W? ----------
dove_df %>%
  group_by(period) %>%
  summarise(records      = n(),
            west_of_100W = sum(longitude < -100),
            percent_west = round(100 * mean(longitude < -100), 1))

✏️ Q10: The dots in panel 1 are a knot in Florida. The hull in panel 1 covers most of the country. How many records does it take to do that?

✏️ Q11: The table says 4.5% of 1990s records were west of 100°W and 65% of recent ones are. The hull panels all look the same. Which of those two things is the better description of a range expansion?

💡 Key idea: a convex hull is set only by its outermost points. Everything inside is ignored, so one escaped pet or one typo’d coordinate can redraw a continent. A hull answers “how far apart are the extremes?” — which is not the same question as “where does this species live?”

🧩 Chunk 3 — Hexagons (after lecture Chunk 3)

Parts 8–10: a grid, a spatial join, and a count.

Part 8 · Build a hex grid

Instead of wrapping the outside, chop the region into equal cells and count the records in each one.

▶ Run this:

# A rectangle over the eastern US ----------------------
east_box <- st_bbox(c(xmin = -92, xmax = -69, ymin = 33, ymax = 45), crs = 4326) %>%
  st_as_sfc()

# Tile it with hexagons --------------------------------
hex_grid <- st_make_grid(east_box, cellsize = 0.5, square = FALSE) %>%
  st_as_sf() %>%
  mutate(hex_id = row_number())

nrow(hex_grid)

st_bbox() builds a bounding box from four numbers; st_as_sfc() turns it into a real shape
square = FALSE is the whole trick — it tiles with hexagons instead of squares
cellsize = 0.5 is half a degree, roughly 55 km across
hex_id gives each cell a name, so we can join to it later

Why hexagons? A square has neighbours at two different distances — edge neighbours are closer than corner ones. A hexagon’s six neighbours are all the same distance away, so spread looks the same in every direction.

✏️ Q12: Change cellsize to 2 and re-run. How many cells now? What would an enormous cell size turn the map into?

Part 9 · Which hexagon is each record in?

▶ Run this:

# A spatial join: point -> the hexagon containing it ----
slf_hex_counts <- st_join(slf_sf, hex_grid) %>%
  st_drop_geometry() %>%
  count(year, hex_id, name = "records")

head(slf_hex_counts)

st_join() is left_join() with geometry as the key. Look at what is missing: there is no by = argument. There does not need to be one — the key is the question “is this point inside that polygon?”, and sf works that out itself.

After that it is count(), the same count() as always: one row per year per occupied cell.

✏️ Q13: st_join() needs no by =. In an ordinary left_join(), what would go wrong if you forgot it?

Part 10 · The hex map, and the number that comes with it

▶ Run this:

# Attach the counts to the hexagons --------------------
slf_hex_map <- hex_grid %>%
  inner_join(slf_hex_counts, by = "hex_id")

# The headline number ----------------------------------
slf_hex_map %>%
  st_drop_geometry() %>%
  count(year, name = "occupied_hexes")

inner_join() is the ordinary join from two weeks ago — and here its dropping behaviour is the point. Cells with no records never match, so they fall out and never get drawn.

▶ Run this:

# The hex map: occupied, or not ------------------------
ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = slf_hex_map, fill = "firebrick", alpha = 0.75, color = NA) +
  coord_sf(xlim = c(-92, -69), ylim = c(33, 45)) +
  facet_wrap(~ year) +
  theme_void()

One flat colour, and that is deliberate. A cell gets drawn if it holds at least one record, so the picture shows exactly what the occupied_hexes table counts — no more, no less.

Now do the same for the dove, on a coarser grid because the region is the whole country.

▶ Run this:

# A grid over the lower 48 -----------------------------
us_box <- st_bbox(c(xmin = -125, xmax = -66, ymin = 24, ymax = 50), crs = 4326) %>%
  st_as_sfc()

us_hex <- st_make_grid(us_box, cellsize = 1.5, square = FALSE) %>%
  st_as_sf() %>%
  mutate(hex_id = row_number())

# Same three steps: join, count, attach ----------------
dove_hex_counts <- st_join(dove_sf, us_hex) %>%
  st_drop_geometry() %>%
  count(period, hex_id, name = "records")

dove_hex_map <- us_hex %>%
  inner_join(dove_hex_counts, by = "hex_id")

dove_hex_map %>%
  st_drop_geometry() %>%
  count(period, name = "occupied_hexes")

▶ Run this:

# The dove in hexagons ---------------------------------
ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = dove_hex_map, fill = "firebrick", alpha = 0.75, color = NA) +
  coord_sf(xlim = c(-125, -66), ylim = c(24, 50)) +
  facet_wrap(~ period, nrow = 1) +
  theme_void()

You still have the records column sitting there unused. Put it on the map and see what happens.

▶ Run this:

# The same cells, shaded by record count ---------------
ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = dove_hex_map, aes(fill = records), color = NA) +
  scale_fill_viridis_b(option = "magma", direction = -1,
                       breaks = c(1, 3, 10, 30, 100)) +
  coord_sf(xlim = c(-125, -66), ylim = c(24, 50)) +
  facet_wrap(~ period, nrow = 1) +
  theme_void()

scale_fill_viridis_b() — the b is for binned. breaks cuts the counts into classes you can read straight off the legend
Try scale_fill_viridis_c() instead, with no breaks. Nearly the whole map goes pale: in the first panel, 104 of the 116 occupied cells hold under 5% of what the busiest cell holds, so a smooth ramp squashes them all into one shade
Now look at where the dark cells are. The darkest hexagon in all four panels is in the first one — 344 records in a single cell over Miami. By the last panel the busiest cell is down to 123, even though the occupied-cell count has gone from 116 to 267

✏️ Q16: The range more than doubles while the busiest cell shrinks. The doves did not get rarer in Miami. What are the dark cells actually a map of — and what does that make the shaded map useful for?

✏️ Q14: The dove’s four hulls were four near-identical blobs. Its four occupied-hex counts are not. Same data, same four panels — why can one summary see the expansion and the other cannot?

✏️ Q15: A single escaped pet dove in Oregon in 1991. What does it do to the hull? What does it do to the hex count?

💡 Key idea: report a range as occupied cells, not hull area, whenever a stray record is possible — which is always. And shade a grid by occupancy when you mean range; shade it by count only when the thing you mean to show is effort.

	Convex hull	Hex grid
Built from	outermost points only	every point
One stray record	redraws the whole range	lights up one cell
Holes in the range	invisible	visible
The number it gives you	area	occupied cells
Lines of code	2	5

🧩 Chunk 4 — Two species, and going interactive (after lecture Chunk 4)

Parts 11–13.

Part 11 · Two species on one map

The two North American flying squirrels — southern (Glaucomys volans, oak and hickory) and northern (Glaucomys sabrinus, spruce and fir) — meet in a narrow band across the upper Midwest and New England. That band runs right over the top of Duluth.

▶ Run this:

# Both species, one file -------------------------------
squirrel_df <- read_csv("data/flying_squirrels_gbif.csv")

count(squirrel_df, species)

squirrel_sf <- squirrel_df %>%
  st_as_sf(coords = c("longitude", "latitude"), crs = 4326)

▶ Run this:

# color = species, and that is the whole trick ---------
ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = squirrel_sf, aes(color = species), size = 0.7, alpha = 0.6) +
  coord_sf(xlim = c(-100, -66), ylim = c(25, 50)) +
  theme_void()

▶ Run this:

# The same data, side by side --------------------------
ggplot() +
  geom_sf(data = states_sf, fill = "grey96", color = "grey80") +
  geom_sf(data = squirrel_sf, aes(color = species), size = 0.7, alpha = 0.6) +
  coord_sf(xlim = c(-100, -66), ylim = c(25, 50)) +
  facet_wrap(~ species) +
  theme_void() +
  theme(legend.position = "none")

⚠️ Watch out! coord_sf() crops this one to the East. The northern flying squirrel has a second population down the Pacific coast; leave the limits off and it sits way out to the left, squashing the part you care about.

✏️ Q17: Which of the two plots lets you find the contact zone, and which lets you see each range on its own? Would you ever make both?

✏️ Q18: There are about 3,400 southern records and 770 northern ones here. Is the southern squirrel four times as common, or four times as findable? (It is nocturnal, it comes to bird feeders, and it lives where more people live.)

✏️ Q19: We did not slice these two by year. With 770 northern records spread over 35 years, what would the panels look like — and what would you be tempted to conclude from them?

Part 12 · Make it interactive with `ggplotly()`

ggplotly() takes a ggplot you have already built and hands back a web version: hover, zoom, pan.

There is one catch on maps. ggplotly() does not understand geom_sf() well — it throws away custom hover text and sometimes forgets your coord_sf() limits. So for the interactive version we drop back to plain longitude and latitude columns, with geom_polygon() for the basemap and geom_point() for the records.

▶ Run this:

# The basemap as an ordinary data frame ----------------
states_df <- map_data("state")

head(states_df)

map_data("state") is the same outlines, handed back as a plain data frame with long, lat, and group. The group column keeps Michigan’s two peninsulas from being joined by a stray line across the lake.

▶ Run this:

# Build an ordinary ggplot -----------------------------
slf_gg <- ggplot() +
  geom_polygon(data = states_df, aes(long, lat, group = group),
               fill = "grey96", color = "grey80") +
  geom_point(data = slf_df,
             aes(longitude, latitude,
                 color = factor(year),
                 text  = paste0(state, " ", year)),
             size = 1, alpha = 0.6) +
  coord_fixed(1.3, xlim = c(-92, -69), ylim = c(33, 45)) +
  theme_void()

# Hand it to plotly ------------------------------------
ggplotly(slf_gg, tooltip = "text")

text = is not a real ggplot aesthetic. ggplot prints a warning and ignores it — that warning is expected
ggplotly() picks it up anyway, and tooltip = "text" says show only that
coord_fixed(1.3) stops the country looking squashed. coord_sf() would handle it properly, but we are off the sf path here on purpose

Hover over a few points. Then box-zoom into eastern Pennsylvania and keep going.

✏️ Q20: Swap tooltip = "text" for tooltip = "all" and hover again. Which is more useful, and why would you ever choose to show less?

Part 13 · A slider instead of an animation

▶ Run this:

# frame = is the only change ---------------------------
slf_anim <- ggplot() +
  geom_polygon(data = states_df, aes(long, lat, group = group),
               fill = "grey96", color = "grey80") +
  geom_point(data = slf_df,
             aes(longitude, latitude,
                 frame = factor(year),
                 text  = paste0(state, " ", year)),
             color = "firebrick", size = 1, alpha = 0.6) +
  coord_fixed(1.3, xlim = c(-92, -69), ylim = c(33, 45)) +
  theme_void()

ggplotly(slf_anim, tooltip = "text")

A play button and a slider appear under the plot. frame = factor(year) is the only line that changed.

This is three scatterplots with a slider on top — not a rendered movie — so it runs on any laptop and needs no extra packages.

✏️ Q21: Compare the slider to the faceted map from Part 4. Which one makes it easier to see how much ground was gained between 2021 and 2025? Which one is more fun? Are those the same thing?

💡 Key idea: interactive is for exploring — you ask it questions by hovering and zooming. Faceted small multiples are for explaining — the comparison is already made for the reader. A figure in a paper is a facet. A figure on a website can be either.

Part 14 · If there is time — was your download honest?

You mapped 1,500 lanternfly records from 2025. GBIF holds about 21,000. The cached file in data/ holds a capped batch from every year, pulled the same way.

▶ Run this:

# The cached file, 2025 only ---------------------------
slf_cached <- read_csv("data/spotted_lanternfly_gbif.csv") %>%
  filter(year == 2025)

# Your live download, 2025 only ------------------------
slf_live <- slf_df %>%
  filter(year == 2025)

# Compare where they say the insect is -----------------
slf_cached %>% count(state, sort = TRUE) %>% head(8)

slf_live %>% count(state, sort = TRUE) %>% head(8)

✏️ Q22: Do the two top-eight lists agree? If they do not, which states appear in one and not the other?

✏️ Q23: Neither list is the truth — both are cutoffs of the same 21,000 records. What would you have to download to get the real answer, and why did we not do that in class?

Homework — a species that moves, of your own

Open 18_mapping_redux_homework.R. That file — not today’s script — is what you turn in.

Pick one species and run today’s recipe on it across time:

Use a cached file — Virginia opossum or giant swallowtail — or download your own species one year at a time
Cut the years into periods with case_when()
Facet the map by period
Build a convex hull and measure its area
Build a hex grid and count occupied hexes per period
Say which of those two numbers you believe, and why
Then the hard one: is your species moving, or are there just more people with phones?

Everything you need is in the homework skeleton, with the lanternfly and dove code sitting right there as worked examples. All the graded questions are in one block at the top of the file — answer them there as # comments.

Other Formats

Activity 18 — Mapping Redux: Ranges That Move

In-class Activity 18: Ranges That Move 🪰🕊️🐿️

Today’s Objectives

🧩 Chunk 1 — Time on a map (after lecture Chunk 1)

Part 1 · Load up and download three snapshot years

Part 2 · Stack them and look hard at the counts

Part 3 · Make it spatial

Part 4 · Put time on the map

🧩 Chunk 2 — From dots to a range (after lecture Chunk 2)

Part 5 · The convex hull

Part 6 · A second species — the Eurasian collared-dove

Part 7 · Now watch the hull fail

🧩 Chunk 3 — Hexagons (after lecture Chunk 3)

Part 8 · Build a hex grid

Part 9 · Which hexagon is each record in?

Part 10 · The hex map, and the number that comes with it

🧩 Chunk 4 — Two species, and going interactive (after lecture Chunk 4)

Part 11 · Two species on one map

Part 12 · Make it interactive with `ggplotly()`

Part 13 · A slider instead of an animation

Part 14 · If there is time — was your download honest?

Homework — a species that moves, of your own

Review and checkpoint

Other Formats

In-class Activity 18: Ranges That Move 🪰🕊️🐿️

Today’s Objectives

🧩 Chunk 1 — Time on a map (after lecture Chunk 1)

Part 1 · Load up and download three snapshot years

Part 2 · Stack them and look hard at the counts

Part 3 · Make it spatial

Part 4 · Put time on the map

🧩 Chunk 2 — From dots to a range (after lecture Chunk 2)

Part 5 · The convex hull

Part 6 · A second species — the Eurasian collared-dove

Part 7 · Now watch the hull fail

🧩 Chunk 3 — Hexagons (after lecture Chunk 3)

Part 8 · Build a hex grid

Part 9 · Which hexagon is each record in?

Part 10 · The hex map, and the number that comes with it

🧩 Chunk 4 — Two species, and going interactive (after lecture Chunk 4)

Part 11 · Two species on one map

Part 12 · Make it interactive with ggplotly()

Part 13 · A slider instead of an animation

Part 14 · If there is time — was your download honest?

Homework — a species that moves, of your own

Review and checkpoint

Part 12 · Make it interactive with `ggplotly()`