Add support for statsbomb api via statsbombpy

[Alek050]

Right now, we do not inherintly support the use of the StatsBomb api, but only those of .json files. This issue will focus on creating a solution for that in the future.

Special thanks to @jan-swiatek for updating the code!

For those who want to use the api right now:

• install statsbombpy: pip install statsbombpy
• import statsbombpy: from statsbombpy import sb
• set credentials: see authentication statsbombpy
• obtain and save the raw files by using fmt="dict":
  For a full

import json
from statsbombpy import sb

competition_id = 9
season_id = 281
match_id = 3895302

with open(f"match.json", "w") as f:
    match = sb.matches(competition_id=competition_id, season_id=season_id, fmt="dict").get(match_id)
    json.dump([match], f)

with open(f"lineups.json", "w") as f:
    lineups = sb.lineups(match_id=match_id, fmt="dict")
    json.dump([v for k, v in lineups.items()], f)

with open(f"events.json", "w") as f:
    events = sb.events(match_id=match_id, fmt="dict")
    json.dump([event for _id, event in events.items()], f)

• Load in the match in databallpy:

from databallpy import get_match

match = get_match(
    event_data_loc="events.json",
    event_match_loc="match_info.json",
    event_lineup_loc="lineup_info.json",
    event_data_provider="statsbomb",
)

[jan-swiatek]

Hi @Alek050, unfortunately this workaround doesn't work any longer.

I've replaced ids with valid ones:

match_info = sb.matches(competition_id=9, season_id=281, fmt="dict")
lineup_info = sb.lineups(match_id=3895052, fmt="dict")
events = sb.events(3895052, fmt="dict")

import json

files = {"match_info": match_info, "lineup_info": lineup_info, "events": events}

for file, data in files.items():
    with open(f"{file}.json", "w") as fp:
        json.dump(data , fp)

from databallpy import get_match

match = get_match(
    event_data_loc="events.json",
    event_match_loc="match_info.json",
    event_lineup_loc="lineup_info.json",
    event_data_provider="statsbomb",
)

But something is wrong with keys:

File ~/Workspace/databallpy/databallpy/utils/get_match.py:150, in get_match(tracking_data_loc, tracking_metadata_loc, event_data_loc, event_metadata_loc, event_match_loc, event_lineup_loc, tracking_data_provider, event_data_provider, check_quality, verbose)
    149 if event_data_loc and event_data_provider:
--> 150     event_data, event_metadata, databallpy_events = load_event_data(
    151         event_data_loc=event_data_loc,
    152         event_metadata_loc=event_metadata_loc,
    153         event_data_provider=event_data_provider,
    154         event_match_loc=event_match_loc,
    155         event_lineup_loc=event_lineup_loc,
    156     )
    157     uses_event_data = True
    159 event_precise_timestamps = {
    160     "opta": True,
    161     "metrica": True,
...
--> 138     match_json = json.load(f)[0]
    140 with open(lineup_loc, "r", encoding="utf-8") as f:
    141     lineup_json = json.load(f)

Do you have any idea what might happened here? Thanks!

[jan-swiatek]

Ok, there is small difference in data structures. Valid approach is:

import json

from statsbombpy import sb

competition_id = 9
season_id = 281
match_id = 3895302

with open(f"match.json", "w") as f:
    match = sb.matches(competition_id=competition_id, season_id=season_id, fmt="dict").get(match_id)
    json.dump([match], f)

with open(f"lineups.json", "w") as f:
    lineups = sb.lineups(match_id=match_id, fmt="dict")
    json.dump([v for k, v in lineups.items()], f)

with open(f"events.json", "w") as f:
    events = sb.events(match_id=match_id, fmt="dict")
    json.dump([event for _id, event in events.items()], f)

This worked for me :)

[Alek050]

Thanks for figuring this out @jan-swiatek! I will update my initial remark so its the first thing others find too!

[jan-swiatek]

@Alek050 I've noticed small error while dumping lineups.json file. It should be ok now :)