README.md

In anthonyshook/nhldata: Pulls NHL Data from the NHL API and

README

nhldata Package

This package can be used to interact with the NHL.com API to pull hockey statistics.

Installation

The package is not on CRAN, but can be installed from github via

install.packages("devtools")
library(devtools)
install_github("anthonyshook/nhldata")

Building a database

The easiest way to interact with the data is to build a local database, and use DBI to connect to it.

The following code will build a brand new database:

# Set up a connection using whatever database flavor you like
# Note, this has been tested with PostgreSQL and SQLite, but should work with MySQL or MariaDB
# Here I'll use an internal function to set up a SQLite database connection for a DB called 'tempdb'
cn <- nhldata::connect_to_db(db_path = '/path/to/database.sqlite')

# Using that connection, we'll build tables with data from the 2016/2017 through 2019/2020 seasons
build_new_nhl_db(seasons = c('20162017', '20172018', '20182019', '20192020'), 
                 num_cores = 8, 
                 conn = cn)

DBI::dbDisconnect(cn)

The /inst/ folder contains two CSVs that describe the database that is build above: nhl_database_dict.csv -- tables, columns, and data-types nhl_database_indexes.csv -- The table names, and indexes that are built during schema construction

Updating an existing database

A wrapper function to update an existing database with new data exists:

# Again, connect to your DB
cn <- nhldata::connect_to_db(db_path = '/path/to/database.sqlite')

# Get all the NHL data for the past 15 days, and upsert it into the Database.
# This will handle duplicates for you, and only insert the new data
nhldata::update_nhl_database(look_back_days = 15, 
                             conn = cn)

Accessing the Data without a database

You can simply pull data into R using fetch functions, but you usually have to build up to it.

# Fetch the games that occurred on a given day (2020-02-02)
games <- fetch_schedule(season = '20192020', start_date = '2020-02-02', end_date = '2020-02-02')

# find the game you want in the table and use the game ID to get data
# Here we'll look at the penguins and capitals
pens_at_caps <- fetch_boxscore_stats('2019020806')

# We can fetch play-by-play information too
pens_at_caps_pbp <- fetch_play_by_play('2019020806')

# At this point we only have IDs for players, so the easiest way to get the link between IDs and players is
# using the fetch_player_info function 
all_skaters <- do.call('rbind', lapply(pens_at_caps$skater_stats$playerid, fetch_player_info))
all_goalies <- do.call('rbind', lapply(pens_at_caps$goalie_stats$playerid, fetch_player_info))

anthonyshook/nhldata documentation built on Feb. 22, 2021, 3:51 a.m.