StatsPerform¶

• Load local event files
  • MA1 and MA3
  • Opta F7 and F24 (or F73)
• Load local tracking files
• Load remote files

Load local event files¶

MA1 and MA3¶

In [ ]:

from kloppy import statsperform

dataset = statsperform.load_event(
    ma1_data="../../kloppy/tests/files/statsperform_event_ma1.json",
    ma3_data="../../kloppy/tests/files/statsperform_event_ma3.json",
    
    # Optional arguments
    coordinates="opta",    
    pitch_length=102.5,
    pitch_width=69.0,
    event_types=["pass", "shot"],
)

dataset.to_df().head()

from kloppy import statsperform dataset = statsperform.load_event( ma1_data="../../kloppy/tests/files/statsperform_event_ma1.json", ma3_data="../../kloppy/tests/files/statsperform_event_ma3.json", # Optional arguments coordinates="opta", pitch_length=102.5, pitch_width=69.0, event_types=["pass", "shot"], ) dataset.to_df().head()

Out[ ]:

	event_id	event_type	result	success	period_id	timestamp	end_timestamp	ball_state	ball_owning_team	team_id	player_id	coordinates_x	coordinates_y	end_coordinates_x	end_coordinates_y	receiver_player_id	set_piece_type	pass_type	body_part_type	is_counter_attack
0	2328589789	PASS	COMPLETE	True	1	0 days 00:00:00.030000	None	alive	75xi6hloabmnjn2kzgj1g8h1s	75xi6hloabmnjn2kzgj1g8h1s	aksjicf4keobpav3tuujngell	50.0	50.0	33.9	51.0	None	KICK_OFF	None	None	None
1	2328589863	PASS	COMPLETE	True	1	0 days 00:00:02.075000	None	alive	75xi6hloabmnjn2kzgj1g8h1s	75xi6hloabmnjn2kzgj1g8h1s	apdrig6xt1hxub1986s3uh1x	33.9	51.0	34.0	88.6	None	None	None	None	None
2	2328589885	PASS	COMPLETE	True	1	0 days 00:00:06.803000	None	alive	75xi6hloabmnjn2kzgj1g8h1s	75xi6hloabmnjn2kzgj1g8h1s	46vr35g415omy60ahbiv40wk5	34.2	88.6	56.1	92.7	None	None	CHIPPED_PASS	None	None
3	2328589929	PASS	COMPLETE	True	1	0 days 00:00:09.373000	None	alive	75xi6hloabmnjn2kzgj1g8h1s	75xi6hloabmnjn2kzgj1g8h1s	2nrmndj0uq3f46c2cb1fbf85	56.8	94.2	48.0	96.6	None	None	None	None	None
4	2328590577	PASS	COMPLETE	True	1	0 days 00:00:40.486000	None	alive	75xi6hloabmnjn2kzgj1g8h1s	75xi6hloabmnjn2kzgj1g8h1s	46vr35g415omy60ahbiv40wk5	46.1	94.7	32.7	73.8	None	FREE_KICK	None	None	None

Opta F7 and F24 (or F73)¶

You can also use the 'old' Opta F7 and F24 format or the F7 and F73 format.

To use F73, simply pass it to the f24_data parameter instead of the f24 file.

In [ ]:

from kloppy import opta

dataset = opta.load(
    f7_data="../../kloppy/tests/files/opta_f7.xml",
    f24_data="../../kloppy/tests/files/opta_f24.xml",
    # f24_data="../../kloppy/tests/files/opta_f73.xml",
    
    # Optional arguments
    coordinates="opta", 
    event_types=["pass", "shot"],
)

dataset.to_df().head()

from kloppy import opta dataset = opta.load( f7_data="../../kloppy/tests/files/opta_f7.xml", f24_data="../../kloppy/tests/files/opta_f24.xml", # f24_data="../../kloppy/tests/files/opta_f73.xml", # Optional arguments coordinates="opta", event_types=["pass", "shot"], ) dataset.to_df().head()

Out[ ]:

	event_id	event_type	result	success	period_id	timestamp	end_timestamp	ball_state	ball_owning_team	team_id	player_id	coordinates_x	coordinates_y	end_coordinates_x	end_coordinates_y	receiver_player_id	set_piece_type	pass_type	body_part_type	is_counter_attack
0	1510681159	PASS	COMPLETE	True	1	0 days 00:00:00.431000	0 days 00:00:03.395000	alive	569	569	48337	50.1	49.4	36.4	45.1	76001	KICK_OFF	None	None	None
1	1646695660	PASS	COMPLETE	True	1	0 days 00:00:03.395000	0 days 00:00:05.299000	alive	569	569	76001	36.4	45.1	28.0	39.8	164266	None	None	None	None
2	1782829017	PASS	COMPLETE	True	1	0 days 00:00:05.299000	0 days 00:00:06.995000	alive	569	569	164266	27.9	39.8	29.1	62.0	77384	None	None	None	None
3	1909884550	PASS	COMPLETE	True	1	0 days 00:00:06.995000	0 days 00:00:08.971000	alive	569	569	77384	29.3	62.6	26.3	37.6	164266	None	None	None	None
4	1515097980	PASS	COMPLETE	True	1	0 days 00:00:08.971000	NaT	alive	569	569	164266	26.3	34.3	29.1	7.4	None	None	None	None	None

Load local tracking files¶

In [ ]:

from kloppy import statsperform

dataset = statsperform.load_tracking(
    ma1_data="../../kloppy/tests/files/statsperform_tracking_ma1.json",
    ma25_data="../../kloppy/tests/files/statsperform_tracking_ma25.txt",

    # Optional arguments
    coordinates="statsperform",
    only_alive=True,
    limit=50,
    sample_rate=(1/2),
    pitch_length=102.5,
    pitch_width=69.0,
    tracking_system="sportvu",
)

dataset.to_df().head()

from kloppy import statsperform dataset = statsperform.load_tracking( ma1_data="../../kloppy/tests/files/statsperform_tracking_ma1.json", ma25_data="../../kloppy/tests/files/statsperform_tracking_ma25.txt", # Optional arguments coordinates="statsperform", only_alive=True, limit=50, sample_rate=(1/2), pitch_length=102.5, pitch_width=69.0, tracking_system="sportvu", ) dataset.to_df().head()

Out[ ]:

	period_id	timestamp	frame_id	ball_state	ball_owning_team_id	ball_x	ball_y	ball_z	ball_speed	a2s2c6anax9wnlsw1s6vunl5h_x	...	6wfwy94p5bm0zv3aku0urfq39_d	6wfwy94p5bm0zv3aku0urfq39_s	6ekdnbnk56xlxforb5owt3dn9_x	6ekdnbnk56xlxforb5owt3dn9_y	6ekdnbnk56xlxforb5owt3dn9_d	6ekdnbnk56xlxforb5owt3dn9_s	ct32113pfx5q9avf2c0x208ru_x	ct32113pfx5q9avf2c0x208ru_y	ct32113pfx5q9avf2c0x208ru_d	ct32113pfx5q9avf2c0x208ru_s
0	1	0 days 00:00:00	1598184000000	alive	None	52.350	33.250	0.0	None	52.803	...	None	None	5.268	33.556	None	None	NaN	NaN	None	None
1	1	0 days 00:00:00.100000	1598184000100	alive	None	50.615	35.325	0.0	None	52.558	...	None	None	5.265	33.529	None	None	NaN	NaN	None	None
2	1	0 days 00:00:00.200000	1598184000200	alive	None	49.630	36.140	0.0	None	52.310	...	None	None	5.264	33.502	None	None	NaN	NaN	None	None
3	1	0 days 00:00:00.300000	1598184000300	alive	None	48.725	36.625	0.0	None	52.059	...	None	None	5.268	33.476	None	None	NaN	NaN	None	None
4	1	0 days 00:00:00.400000	1598184000400	alive	None	47.890	37.130	0.0	None	51.804	...	None	None	5.277	33.452	None	None	NaN	NaN	None	None

5 rows × 101 columns

In [ ]:

from kloppy import statsperform

dataset = statsperform.load_tracking(
    ma1_data="../../kloppy/tests/files/statsperform_tracking_ma1.xml",
    ma25_data="../../kloppy/tests/files/statsperform_tracking_ma25.txt",

    # Optional arguments
    coordinates="statsperform",
    only_alive=True,
    limit=50,
    sample_rate=(1/2),
    pitch_length=102.5,
    pitch_width=69.0,
    tracking_system="sportvu",
)

dataset.to_df().head()

from kloppy import statsperform dataset = statsperform.load_tracking( ma1_data="../../kloppy/tests/files/statsperform_tracking_ma1.xml", ma25_data="../../kloppy/tests/files/statsperform_tracking_ma25.txt", # Optional arguments coordinates="statsperform", only_alive=True, limit=50, sample_rate=(1/2), pitch_length=102.5, pitch_width=69.0, tracking_system="sportvu", ) dataset.to_df().head()

Out[ ]:

	period_id	timestamp	frame_id	ball_state	ball_owning_team_id	ball_x	ball_y	ball_z	ball_speed	a2s2c6anax9wnlsw1s6vunl5h_x	...	6wfwy94p5bm0zv3aku0urfq39_d	6wfwy94p5bm0zv3aku0urfq39_s	6ekdnbnk56xlxforb5owt3dn9_x	6ekdnbnk56xlxforb5owt3dn9_y	6ekdnbnk56xlxforb5owt3dn9_d	6ekdnbnk56xlxforb5owt3dn9_s	ct32113pfx5q9avf2c0x208ru_x	ct32113pfx5q9avf2c0x208ru_y	ct32113pfx5q9avf2c0x208ru_d	ct32113pfx5q9avf2c0x208ru_s
0	1	0 days 00:00:00	1598184000000	alive	None	52.350	33.250	0.0	None	52.803	...	None	None	5.268	33.556	None	None	NaN	NaN	None	None
1	1	0 days 00:00:00.100000	1598184000100	alive	None	50.615	35.325	0.0	None	52.558	...	None	None	5.265	33.529	None	None	NaN	NaN	None	None
2	1	0 days 00:00:00.200000	1598184000200	alive	None	49.630	36.140	0.0	None	52.310	...	None	None	5.264	33.502	None	None	NaN	NaN	None	None
3	1	0 days 00:00:00.300000	1598184000300	alive	None	48.725	36.625	0.0	None	52.059	...	None	None	5.268	33.476	None	None	NaN	NaN	None	None
4	1	0 days 00:00:00.400000	1598184000400	alive	None	47.890	37.130	0.0	None	51.804	...	None	None	5.277	33.452	None	None	NaN	NaN	None	None

5 rows × 101 columns

Load remote files¶

Kloppy supports remote files through fsspec FileSystem under the hood. This allows you to work with files in AWS S3, Google Cloud, Azure Blob, HDFS, FTP, and SFTP without extra tools. For example you can pass:

• Individual s3 file paths: (e.g ma1_data=s3://.../statsperform_event_ma1.xml)

Note: Kloppy might throw an the first time to help you identify missing cloud specific dependencies like s3fs.

In [ ]:

from kloppy import statsbomb

dataset = statsperform.load_event(
    ma1_data="s3://.../statsperform_event_ma1.xml",
    ma3_data="s3://.../statsperform_event_ma3.xml",
    
    # Optional arguments
    coordinates="opta",
    pitch_length=102.5,
    pitch_width=69.0,
    event_types=["pass", "shot"],
)

from kloppy import statsbomb dataset = statsperform.load_event( ma1_data="s3://.../statsperform_event_ma1.xml", ma3_data="s3://.../statsperform_event_ma3.xml", # Optional arguments coordinates="opta", pitch_length=102.5, pitch_width=69.0, event_types=["pass", "shot"], )