fix: event_num typo, missing DB write in load_nba_team_box, uninitialized games in scoreboards

[jacobpstein]

Summary

I'm a big fan of hoopR and use it extensively for my Wizards coverage at wizardspoints.substack.com. While working with the package I ran into a few bugs I've been meaning to address — fixing them here.

Changes

• R/nba_stats_pbp.R (lines 154–155): Typo even_num → event_num in the .players_on_court() fallback branch (V2 path). The correct column name event_num is used at lines 99–100 in the same function; this branch missed the update. Triggers when a player is subbed on and off at the exact same clock time in the same period.

• R/load_nba.R: load_nba_team_box() accepts dbConnection and tablename parameters and sets in_db <- TRUE, but never applies the flag. The if (in_db) DBI::dbWriteTable(...) block present in load_nba_pbp() and load_nba_player_box() was missing entirely. Callers passing a DB connection silently received a tibble instead of writing to the database.

• R/nba_stats_scoreboard.R: nba_scoreboardv3(), nba_schedule(), and nba_todays_scoreboard() assign games inside their tryCatch expr block but call return(games) outside it. If the API errors, games is never defined and the function crashes with "object 'games' not found" instead of returning gracefully. Added games <- list() before each tryCatch, matching the pattern used by the deprecated nba_scoreboardv2() and all other df_list-returning functions in the file.

Test plan

• nba_pbp(game_id = "0022200021", version = "v2", on_court = TRUE) — confirm no column-not-found error
• load_nba_team_box(seasons = 2023, dbConnection = con, tablename = "team_box") with a DBI connection — confirm data written to DB
• Call nba_scoreboardv3(), nba_schedule(), nba_todays_scoreboard() with an invalid date — confirm graceful list() return instead of crash
• devtools::check() — 0 new errors or warnings

Summary by CodeRabbit

• New Features

  • Team box score data can now be persisted to a configured database for easier data management

• Bug Fixes

  • Fixed substitution event tracking in play-by-play data to ensure accurate player rotation information
  • Enhanced robustness of scoreboard query functions during error conditions

[vercel]

Someone is attempting to deploy a commit to the sportsdataverse Team on Vercel.

A member of the Team first needs to authorize it.

[coderabbitai]

📝 Walkthrough

Walkthrough

These changes modify three R functions: adding conditional database writing to load_nba_team_box(), correcting a column reference in .players_on_court(), and explicitly initializing a variable in three scoreboard functions before error handling blocks.

Changes

Cohort / File(s)	Summary
Database Persistence R/load_nba.R	Modified load_nba_team_box() to conditionally write loaded data to database when in_db is true and dbConnection/tablename are provided. Returns NULL on database write; otherwise returns in-memory classed tibble as before.
Play-by-Play Data Processing R/nba_stats_pbp.R	Fixed column reference in .players_on_court() substitution handling: changed from pbp_data_period$even_num to pbp_data_period$event_num for proper event ordering comparison.
Scoreboard API Error Handling R/nba_stats_scoreboard.R	Explicitly initialized games <- list() before tryCatch() blocks in nba_schedule(), nba_scoreboardv3(), and nba_todays_scoreboard() to ensure defined return value on error paths.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Poem

🐰 A bunny hops through data with care,
Writes to databases, no column repair!
Event_num fixed, errors now caught,
Clean initialization—just what we sought! 🎲

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately summarizes all three main bug fixes in the changeset: the event_num typo in nba_stats_pbp.R, the missing DB write in load_nba_team_box, and uninitialized games in scoreboards functions.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

🧹 Nitpick comments (1)

R/nba_stats_scoreboard.R (1)

88-88: LGTM — prevents object 'games' not found on API error.

Initializing games <- list() before each tryCatch ensures the error path returns a valid (empty) value instead of crashing. One minor note: on the success path these functions return a tibble while the error path now returns list() — the return type is inconsistent across branches. If callers downstream expect a data frame shape, consider games <- dplyr::tibble() instead. Non-blocking; the pattern matches existing sibling functions (df_list <- list()).

Also applies to: 684-684, 837-837

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@R/nba_stats_scoreboard.R` at line 88, You initialized games <- list() to
avoid "object 'games' not found" on API error, but that makes the error path
return a list while the success path returns a tibble; change the initializer to
dplyr::tibble() (or tibble::tibble()) so both success and error branches return
the same data-frame-like type, and apply the same change to the other
occurrences of games <- list() in the file (the similar blocks around the other
tryCatch usages).

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@R/nba_stats_scoreboard.R`:
- Line 88: You initialized games <- list() to avoid "object 'games' not found"
on API error, but that makes the error path return a list while the success path
returns a tibble; change the initializer to dplyr::tibble() (or
tibble::tibble()) so both success and error branches return the same
data-frame-like type, and apply the same change to the other occurrences of
games <- list() in the file (the similar blocks around the other tryCatch
usages).

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: db470937-c62e-491a-ac48-99b02c2a5f8c

📥 Commits

Reviewing files that changed from the base of the PR and between d425ec5 and 3153114.

📒 Files selected for processing (3)

• R/load_nba.R
• R/nba_stats_pbp.R
• R/nba_stats_scoreboard.R