Skip to content

A Python SDK for the Yahoo! Fantasy Sports API

Notifications You must be signed in to change notification settings

jackadair/yahoofantasy-nhl

 
 

Repository files navigation

🏆 Yahoo Fantasy API Wrapper 🏆

The Yahoo Fantasy Sports API is difficult to comprehend, has this strange one-page documentation setup that is hard to navigate, and seems to only want to conform to a small portion of the OAuth spec. This library/SDK makes your life easier if you want to write an app that interfaces with the Yahoo Fantasy Sports API.

This library will, in theory, work for any Yahoo Fantasy Sports API leagues/teams. It contains some common constructs and helper methods for head-to-head leagues and the sports of NFL 🏈, MLB ⚾, and NBA 🏀 at this time. More sports and league types are planned for the future.

Table of Contents

Installation

pip install yahoofantasy

You will also need a application registered on the Yahoo Developer Site. You'll need your client ID and secret. The app just needs to have read permissions. See below for instructions on how to set up your Yahoo Developer application if you don't have one already.

Basic Usage

You're going to want to start off by logging in to your Yahoo Developer application, then creating a context. This context is where all of your API requests will originate and league information will live.

yahoofantasy login

Once you've logged in, create a context and use that to make requests. For example, to fetch all of your leagues for a given game/season:

from yahoofantasy import Context

ctx = Context()
leagues = ctx.get_leagues('mlb', 2020)
for league in leagues:
    print(league.name + " -- " + league.league_type)

Authentication

You can use the built-in yahoofantasy CLI to obtain an access token and refresh token for your application. Follow these steps:

  1. Set up your Yahoo application to have a callback/redirect URI of https://localhost:8000. If you already have an app that points to your local host on a different port or different path that's ok, you can customize later on.
  2. Install yahoofantasy if you haven't already
pip install yahoofantasy
  1. Log in with your Yahoo account. This command will launch a browser that will ask you to authenticate to your app. It will then store the token in a local file that can be consumed by the yahoofantasy SDK.
yahoofantasy login

NOTE: If you see a browser certificate warning that is ok, proceed anyway past the warning to save your token. This warning happens because Yahoo requires an HTTPS redirect URI and the local server uses an untrusted certificate.

Try yahoofantasy login --help for some advanced options, like customizing the port or redirect URI

Concepts

There is a general hierarchy that head-to-head leagues will follow. This hierarchy is represented with classes within this library. This code walkthrough will help you understand the organization of the library. The following examples are intended to be read sequentially and assume you have a Context with your logged in Yahoo credentials called ctx.

  • Your account will belong to one or more League objects.
for league in ctx.get_leagues('mlb', 2019):
    print(f"{league.id} - {league.name} ({league.league_type})")
  • A League will contain multiple Team objects.
from yahoofantasy import League

league = League(ctx, '388.l.25000')  # Use a manual league ID or get it from league.id above
for team in league.teams():
    print(f"Team Name: {team.name}\tManager: {team.manager.nickname}")
  • A Team has multiple Player objects that define their lineup. This is their current lineup and not a lineup for a given week
for team in league.teams():
    players = team.players()
    for player in players:
        print(f"Player: {player.name.full}")
  • A League has Standings, which is an ordered list of Team objects.
for team in league.standings():
    outcomes = team.team_standings.outcome_totals
    print(f"#{team.team_standings.rank}\t{team.name}\t"
          f"({outcomes.wins}-{outcomes.losses}-{outcomes.ties})")
  • A League will contain multiple Week objects. A Week contains multiple Matchup objects, which are a head-to-head matchup of two Team objects for that week.
week_3 = league.weeks()[2]
for matchup in week_3.matchups:
    print(f"{matchup.team1.name} vs {matchup.team2.name}")
  • A Matchup will have multiple Stat objects for the two teams. A Stat object contains the display name of the stat as well as the value for the team.
matchup = week_3.matchups[0]
print(f"{matchup.team1.name}\tvs\t{matchup.team2.name}")
for team1_stat, team2_stat in zip(matchup.team1_stats, matchup.team2_stats):
    print(f"{team1_stat.value}\t{team1_stat.display}\t{team2_stat.value}")

The full sequence of these examples can be run in the examples folder under the readme.py script, like so:

cd examples
yahoofantasy login
python readme.py

Command Line (CLI)

This package comes with a built in CLI to let you do some handy tasks without writing any Python code. This is useful for exporting a spreadsheet with trades in your league, player performances, etc and doing some separate analysis on them.

General Properties

Each CLI command has these common properties/arguments to let you control its behavior

  • -g/--game - which sport you are exporting (e.g., nfl, mlb)
  • -s/--season - which season you are exporting (e.g., 2020, 2019, etc)
  • -o/--output - the filename of the CSV to write to, defaults to stdout which prints to stdout instead of to a file

If you don't provide these parameters you will be prompted for the required ones when you run your command. These parameters must be provided after the dump command but before the type of export you want to complete. For example:

yahoofantasy dump -g nfl -s 2020 -o path/to/output.csv performances

Types of Exports

Player Performances

Dumps a CSV with every player that was owned for every week and their stats.

yahoofantasy dump performances

Simplified output example:

name week manager position points Pass TD Rush Yds
Drew Brees 1 Manager Name QB 16.4 2 2
Dalvin Cook 1 Manager RB 21.3 0 50

Matchups

In a head-to-head league, a CSV dump of all manager matchups from the season

yahoofantasy dump matchups

Simplified output example:

week manager win points proj_points opponent opp_points opp_proj_points
1 Manager 1 False 90.0 133.55 Manager 2 142.68 136.79
1 Manager 2 True 142.68 136.79 Manager 1 90.0 133.55

Draft Results

A CSV dump of every draft pick

yahoofantasy dump draftresults

Simplified output example:

pick round manager player pos
1 1 Manager 1 Christian McCaffrey RB
2 1 Manager 2 Saquon Barkley RB

Transactions

A CSV dump of every transaction made for a season. Includes trades, adds, drops, and commissioner moves

yahoofantasy dump transactions

Simplified output example:

type player_type player from to ts week_idx bid
drop drop Damien Harris Elementary Mr Watson waivers 09/09/2020, 20:57:15 36
add/drop add James Robinson waivers Kittles taste the 🌈 09/11/2020, 00:22:20 36

Shell (Python interpreter)

A command is available to drop you into a Python interpreter with access to your Context object as the ctx variable. From the directory where you ran yahoofantasy login you can run:

yahoofantasy shell

Clearing cache

The yahoofantasy library maintains its own persisted cache of certain Yahoo! API responses. This cuts down on the number of requests that need to be made and makes future function calls faster. Occasionally you may want to clear this cache but not lose your authentication data. You can do so by running the following CLI command from your yahoofantasy project directory:

yahoofantasy clear-cache

Development

Issues, pull requests, and contributions are more than welcome.

Unit Tests

To run the tests, after install:

py.test

Or to keep running tests using testmon and drop into a pdb shell on failure (my preferred mode):

pytest-watch --pdb -- --testmon -s

Releasing

I use bump2version to manage version bumping. This will update the version number in the library, commit it, and create a version tag.

bump2version minor
git push

About

A Python SDK for the Yahoo! Fantasy Sports API

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%