Fix plan: cow column bugs across Munro and Corbett blog posts

This plan describes all the changes needed to fix the cow-related bugs in the two blog posts and their supporting scripts. It is intended to be executed by Sonnet in order, and each step includes the exact file, the problem, and the fix.

Context

There are two blog posts and two scraping scripts involved:

Munro post: post/2025-08-31-munro-tidy-tuesday/index.en.qmd
Munro scrape: post/2025-08-31-munro-tidy-tuesday/scraping script.R
Corbett post: post/2025-09-25-corbetts/corbetts.qmd
Corbett scrape: post/2025-09-25-corbetts/corbett_scraping.R
Patch file (to delete): post/patch_cow_column.R

Both scraping scripts already include cow in extract_flags_time_distance(), so a fresh scrape will produce CSVs with a native cow column. The patch file is therefore redundant.

Step 1: Re-run the Munro scrape

Run post/2025-08-31-munro-tidy-tuesday/scraping script.R in R. This will regenerate walkhighlands.csv with the cow column included natively. This takes a while because of the polite delays between HTTP requests (~0.2-0.6s per page, ~282 Munros plus rating/ascent pages).

No code changes are needed to the scraping script itself. It already scrapes cow correctly.

Step 2: Re-run the Corbett scrape

Run post/2025-09-25-corbetts/corbett_scraping.R in R. This will regenerate walkhighlands_corbetts.csv with the cow column included natively. Same timing considerations as Step 1 (~220 Corbetts).

Important: The Corbett scraping script writes output to the current working directory (write_csv(walkhighlands_corbetts, file = "walkhighlands_corbetts.csv") at line 251). Make sure R’s working directory is set to post/2025-09-25-corbetts/ before running, or change line 251 to use the full path. Alternatively, after running, copy the output CSV to post/2025-09-25-corbetts/walkhighlands_corbetts.csv.

Step 3: Fix the Munro QMD cow plot (Bug 1 — wrong column name)

File: post/2025-08-31-munro-tidy-tuesday/index.en.qmd
Line: 944
Problem: filter(munros_coords, cows == TRUE) — the column is called cow, not cows.
Fix: Change cows to cow.

Before:

  geom_jitter(data = filter(munros_coords, cows == TRUE),

After:

  geom_jitter(data = filter(munros_coords, cow == TRUE),

Step 4: Fix the Corbett QMD cow plot (Bug 2 — wrong data object AND wrong column name)

File: post/2025-09-25-corbetts/corbetts.qmd
Lines: 907-909
Problem: Two errors on this line. It references munros_coords (which does not exist in the Corbett environment) and uses cows (the column is cow).
Fix: Change munros_coords to corbett_coords and cows to cow.

Before:

  geom_jitter(data = filter(munros_coords, cows == TRUE),
             aes(x = longitude,
                 y = latitude,
                 colour = cow,

After:

  geom_jitter(data = filter(corbett_coords, cow == TRUE),
             aes(x = longitude,
                 y = latitude,
                 colour = cow,

Step 5: Remove the defensive cow fallback from the Munro QMD

File: post/2025-08-31-munro-tidy-tuesday/index.en.qmd
Lines: 225-228
Problem: This code exists as a safety net for when walkhighlands.csv did not have a cow column. After re-scraping (Step 1), the column will always be present, so this code is unnecessary and masks potential data issues.

Delete these lines:

# Ensure cow column exists and has no NAs.
# The column is absent if walkhighlands.csv pre-dates patch_cow_column.R.
if (!"cow" %in% names(munro_dat)) munro_dat$cow <- FALSE
munro_dat <- munro_dat |> mutate(cow = ifelse(is.na(cow), FALSE, cow))

Step 6: Remove the defensive cow fallback from the Corbett QMD

File: post/2025-09-25-corbetts/corbetts.qmd
Lines: 176-179
Problem: Same as Step 5 but for the Corbett post.

Delete these lines:

# Ensure cow column exists and has no NAs.
# The column is absent if walkhighlands_corbetts.csv pre-dates patch_cow_column.R.
if (!"cow" %in% names(corbett_dat)) corbett_dat$cow <- FALSE
corbett_dat <- corbett_dat |> mutate(cow = ifelse(is.na(cow), FALSE, cow))

Step 7: Delete the patch file

File: post/patch_cow_column.R
Action: Delete this file entirely. It is no longer needed since both scraping scripts produce the cow column natively.

Step 8: Verify the Corbett QMD cow section text says “toggle the graph”

File: post/2025-09-25-corbetts/corbetts.qmd
Line: 901
Note: The text reads “You can toggle the graph” which is a leftover fragment that does not make sense in context (it is a static plotly map, not a toggleable chart). Consider removing this sentence or replacing it with something more appropriate. This is a minor editorial note, not a bug.

Step 9: Re-render both QMDs

After all changes are made:

Open post/2025-08-31-munro-tidy-tuesday/index.en.qmd in RStudio and render it. Confirm the cow map appears correctly under “Other features > Cows”.
Open post/2025-09-25-corbetts/corbetts.qmd in RStudio and render it. Confirm the cow map appears correctly under “Cows”.

Summary of changes

File	Action
`scraping script.R`	Re-run (no code changes)
`corbett_scraping.R`	Re-run (no code changes)
`index.en.qmd` line 944	`cows` → `cow`
`index.en.qmd` lines 225-228	Delete defensive fallback
`corbetts.qmd` line 907	`munros_coords` → `corbett_coords`, `cows` → `cow`
`corbetts.qmd` lines 176-179	Delete defensive fallback
`corbetts.qmd` line 901	Remove “You can toggle the graph” (editorial)
`patch_cow_column.R`	Delete file