[Football Analysis] 15.활동량 지표 집계

작성일자 : 2024-05-08

Ver 0.1.1

 

데이터 준비¶

In [1]:

import os

# directory 변경
new_dir = '/Users/limjongjun/Desktop/JayJay/Growth/Python/soccer-analytics'
os.chdir(new_dir)

import numpy as np
import pandas as pd
from tqdm import tqdm
import matplotlib.pyplot as plt
from src.plot_utils import draw_pitch

(1) 가공 데이터 불러오기¶

In [3]:

match_id = 1
file = f'data_metrica/data/Sample_Game_{match_id}/Sample_Game_{match_id}_IntegratedData.csv'
traces = pd.read_csv(file, header=0, index_col=0)
traces

Out[3]:

	period	time	H11_x	H11_y	H01_x	H01_y	H02_x	H02_y	H03_x	H03_y	...	A24_speed	A26_vx	A26_vy	A26_speed	A27_vx	A27_vy	A27_speed	A28_vx	A28_vy	A28_speed
frame
1	1	0.04	0.08528	32.80184	33.95392	44.41896	35.04904	33.22684	32.16408	24.15972	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	1	0.08	0.09984	32.80184	33.95392	44.41896	35.04904	33.22684	32.16408	24.15972	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
3	1	0.12	0.11856	32.80184	33.95392	44.41896	35.04904	33.22684	32.16408	24.15972	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
4	1	0.16	0.12584	32.80184	33.92688	44.41556	35.03448	33.31184	32.18176	24.17672	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
5	1	0.20	0.13416	32.80184	33.90088	44.38292	35.01056	33.33224	32.18592	24.15904	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
145002	2	5800.08	9.76144	37.15860	NaN	NaN	68.47672	24.07336	71.52288	22.09116	...	0.499132	0.354571	0.405385	0.538570	0.298571	-0.494308	0.577482	0.0	0.0	0.0
145003	2	5800.12	9.82800	37.15860	NaN	NaN	68.50792	24.08356	71.52080	22.08028	...	0.499132	0.453857	0.299462	0.543750	0.359857	-0.606769	0.705455	0.0	0.0	0.0
145004	2	5800.16	9.88832	37.15860	NaN	NaN	68.54744	24.09308	71.53744	22.09932	...	0.499132	0.553143	0.193538	0.586024	0.421143	-0.719231	0.833459	0.0	0.0	0.0
145005	2	5800.20	9.92576	37.15792	NaN	NaN	68.58176	24.10464	71.58216	22.12176	...	0.499132	0.652429	0.087615	0.658285	0.482429	-0.831692	0.961483	0.0	0.0	0.0
145006	2	5800.24	9.92576	37.15792	NaN	NaN	68.64416	24.11008	71.63312	22.07824	...	0.499132	0.751714	-0.018308	0.751937	0.543714	-0.944154	1.089519	0.0	0.0	0.0

145006 rows × 147 columns

• 컬럼별로 선수별 x,y 가 존재하기 때문에 선수별로 데이터를 집계하기에는 어려운 상태

(2) 가공 데이터 형태 변환¶

In [4]:

players = [c[:-2] for c in traces.columns if c.endswith('_x') and not c.startswith('ball')]
player_cols = ['period', 'time', 'x', 'y', 'vx', 'vy', 'speed']
trace_list = []

for p in players:
    player_trace = traces[['period', 'time', f'{p}_x', f'{p}_y', f'{p}_vx', f'{p}_vy', f'{p}_speed']].copy()
    player_trace['team'] = 'Home' if p[0] == 'H' else 'Away'
    player_trace['player_id'] = int(p[1:])
    player_trace.columns = player_cols + ['team', 'player_id']
    trace_list.append(player_trace)

traces = pd.concat(trace_list).reset_index()[['team', 'player_id', 'frame'] + player_cols]
traces

Out[4]:

	team	player_id	frame	period	time	x	y	vx	vy	speed
0	Home	11	1	1	0.04	0.08528	32.80184	NaN	NaN	NaN
1	Home	11	2	1	0.08	0.09984	32.80184	NaN	NaN	NaN
2	Home	11	3	1	0.12	0.11856	32.80184	NaN	NaN	NaN
3	Home	11	4	1	0.16	0.12584	32.80184	NaN	NaN	NaN
4	Home	11	5	1	0.20	0.13416	32.80184	NaN	NaN	NaN
...	...	...	...	...	...	...	...	...	...	...
4060163	Away	28	145002	2	5800.08	81.33008	12.86628	0.0	0.0	0.0
4060164	Away	28	145003	2	5800.12	81.33008	12.86628	0.0	0.0	0.0
4060165	Away	28	145004	2	5800.16	81.33008	12.86628	0.0	0.0	0.0
4060166	Away	28	145005	2	5800.20	81.33008	12.86628	0.0	0.0	0.0
4060167	Away	28	145006	2	5800.24	81.33008	12.86628	0.0	0.0	0.0

4060168 rows × 10 columns

• 뛴 거리는 각 지점별 최단 거리의 합들로 계산

In [5]:

traces['distance'] = traces['speed'] * 0.04
traces['speed'] = traces['speed'] * 3.6 # m/s -> km/h 로 변경
traces.to_csv('data_metrica/data/Sample_Game_1/Sample_Game_1_IntegratedData_Reshaped.csv')
traces[34700:34720]

Out[5]:

	team	player_id	frame	period	time	x	y	vx	vy	speed	distance
34700	Home	11	34701	1	1388.04	4.92752	33.32612	4.948	-1.671231	18.801417	0.208905
34701	Home	11	34702	1	1388.08	5.14696	33.22820	5.042	-1.805923	19.280385	0.214227
34702	Home	11	34703	1	1388.12	5.36536	33.13028	5.140	-1.938000	19.775585	0.219729
34703	Home	11	34704	1	1388.16	5.58480	33.03168	5.238	-2.072692	20.279442	0.225327
34704	Home	11	34705	1	1388.20	5.80528	32.93308	5.288	-2.399615	20.905153	0.232279
34705	Home	11	34706	1	1388.24	6.02680	32.83380	5.196	-2.828538	21.297597	0.236640
34706	Home	11	34707	1	1388.28	6.24832	32.73520	4.840	-3.439231	21.375001	0.237500
34707	Home	11	34708	1	1388.32	6.46984	32.63524	4.554	-3.555615	20.799564	0.231106
34708	Home	11	34709	1	1388.36	6.69240	32.53596	4.432	-3.503308	20.337862	0.225976
34709	Home	11	34710	1	1388.40	6.91600	32.43600	4.426	-3.240462	19.747589	0.219418
34710	Home	11	34711	1	1388.44	7.11464	32.23608	4.478	-2.892615	19.191650	0.213241
34711	Home	11	34712	1	1388.48	7.23944	31.98312	4.554	-2.552615	18.794193	0.208824
34712	Home	11	34713	1	1388.52	7.22696	31.63496	4.640	-2.196923	18.481739	0.205353
34713	Home	11	34714	1	1388.56	7.29560	31.47720	4.684	-1.873923	18.161793	0.201798
34714	Home	11	34715	1	1388.60	7.45160	31.40648	4.626	-1.616308	17.640855	0.196009
34715	Home	11	34716	1	1388.64	7.66688	31.44524	4.470	-1.401846	16.864789	0.187387
34716	Home	11	34717	1	1388.68	7.91336	31.52752	4.224	-1.264538	15.873197	0.176369
34717	Home	11	34718	1	1388.72	8.17336	31.60572	4.004	-0.928462	14.796857	0.164410
34718	Home	11	34719	1	1388.76	8.43960	31.69140	3.878	-0.512615	14.082240	0.156469
34719	Home	11	34720	1	1388.80	8.68400	31.76076	3.982	0.082385	14.338268	0.159314

활동량 지표 집계¶

(1) 뛴 거리¶

In [6]:

distances = traces.groupby('player_id')['distance'].sum()
distances

Out[6]:

player_id
1      3086.510866
2      9774.328865
3      9498.537998
4     10184.797706
5     11006.797770
6      8661.772918
7     10543.077091
8     11119.621735
9     10066.199448
10     9080.472763
11     3696.155770
12     7320.100725
13     2439.279263
14     1925.185387
15     9770.806183
16     9836.752563
17    10849.339113
18    11165.790709
19    10746.203857
20    10218.125707
21    10893.032354
22     9283.927933
23    10222.688175
24     8620.446836
25     4867.230563
26     2788.394785
27     2844.054805
28     2107.814431
Name: distance, dtype: float64

(2) 출전 시간¶

In [7]:

durations = traces[['player_id', 'x']].dropna().groupby('player_id').count() * 0.04 # frame 1개당 0.04 초
durations.columns = ['duration']
durations

Out[7]:

	duration
player_id
1	1868.24
2	5800.24
3	5800.24
4	5800.24
5	5800.24
6	4433.96
7	5800.24
8	5800.24
9	5800.24
10	4855.68
11	5800.24
12	3932.04
13	1366.32
14	944.60
15	5800.24
16	5800.24
17	5800.24
18	5800.24
19	4775.24
20	5800.24
21	5800.24
22	4274.76
23	5800.24
24	4263.96
25	5800.24
26	1536.32
27	1525.52
28	1025.04

(3) 1분당 뛴 거리¶

• 선수별로 출전시간이 다르기 때문에 1분당 뛴 거리로 scailing

In [8]:

stats = pd.concat([durations, distances], axis=1)
stats['dist_1min'] = stats['distance'] / stats['duration'] * 60
stats

Out[8]:

	duration	distance	dist_1min
player_id
1	1868.24	3086.510866	99.125729
2	5800.24	9774.328865	101.109563
3	5800.24	9498.537998	98.256672
4	5800.24	10184.797706	105.355617
5	5800.24	11006.797770	113.858714
6	4433.96	8661.772918	117.210434
7	5800.24	10543.077091	109.061802
8	5800.24	11119.621735	115.025810
9	5800.24	10066.199448	104.128789
10	4855.68	9080.472763	112.204339
11	5800.24	3696.155770	38.234512
12	3932.04	7320.100725	111.699282
13	1366.32	2439.279263	107.117480
14	944.60	1925.185387	122.285754
15	5800.24	9770.806183	101.073123
16	5800.24	9836.752563	101.755299
17	5800.24	10849.339113	112.229899
18	5800.24	11165.790709	115.503400
19	4775.24	10746.203857	135.024047
20	5800.24	10218.125707	105.700375
21	5800.24	10893.032354	112.681879
22	4274.76	9283.927933	130.308058
23	5800.24	10222.688175	105.747571
24	4263.96	8620.446836	121.301985
25	5800.24	4867.230563	50.348578
26	1536.32	2788.394785	108.898984
27	1525.52	2844.054805	111.859096
28	1025.04	2107.814431	123.379445

(4) 속도 구간별 뛴 거리¶

In [9]:

# FIFA에서 나누는 zone 1 ~ 5 기준점이 되는 속도
speed_bins = [0, 7, 15, 20, 25, 50] 
speed_cats = pd.cut(traces['speed'], speed_bins, right=False, labels=np.arange(1, 6))
distances_by_speed = traces.pivot_table('distance', index='player_id', columns=speed_cats, aggfunc='sum')
distances_by_speed

Out[9]:

speed	1	2	3	4	5
player_id
1	1217.931550	1282.799118	455.477781	120.126291	10.176127
2	3725.625280	4825.400329	927.748290	212.003771	83.551196
3	4081.893197	4252.977082	878.452741	171.691357	113.523620
4	3687.151282	4680.773126	1067.728191	536.396559	212.748548
5	3847.243775	5015.489516	1130.576320	647.675724	365.812436
6	2766.991173	4328.730604	1307.106794	232.435299	26.509048
7	3885.758813	4575.688296	1506.997724	490.907722	83.724537
8	3898.154833	4464.576597	1522.565696	783.532281	450.792328
9	3951.717889	3569.714419	1467.116554	768.268454	309.382132
10	3910.213073	3267.428039	1094.716599	607.130206	200.984845
11	3282.563229	371.176295	33.802253	6.238904	2.375089
12	2511.083830	3625.267429	980.233156	192.127237	11.389073
13	945.750675	889.513095	279.182292	224.779260	100.053942
14	564.131943	775.154661	351.832959	224.012757	10.053068
15	4132.798862	4206.381171	935.262943	336.529421	159.833786
16	4024.257560	4247.959630	1016.137160	391.781403	156.616809
17	4147.272702	4247.127041	1774.457437	546.592458	133.889475
18	3526.730228	5295.385219	1564.463895	535.277400	243.933967
19	3677.541130	4506.565476	2028.468939	453.380768	80.247543
20	3395.837437	4754.236541	1340.282890	567.465522	160.303316
21	3499.442314	4977.847699	1769.059185	501.451468	145.231689
22	4505.327415	3048.550862	1047.319116	521.782919	160.947621
23	3963.602878	4291.647139	1110.694186	440.797862	415.946109
24	3280.351892	3463.719445	1309.653142	445.626703	121.095654
25	3852.077039	890.742902	108.892115	15.518508	0.000000
26	999.136621	1233.878420	374.667924	167.182760	13.529060
27	930.092231	1222.834657	472.294892	173.719479	45.113546
28	620.928677	851.555307	361.364594	231.876492	42.089360

(5) 추가 지표 산출 및 연결¶

In [10]:

distances_by_speed.columns = [f'zone{i}_dist' for i in distances_by_speed.columns]
stats = pd.concat([stats, distances_by_speed], axis=1)

grouped = traces.groupby('player_id')
stats['max_speed'] = grouped['speed'].max()
stats['mean_x'] = grouped['x'].mean()
stats['mean_y'] = grouped['y'].mean()
stats['team'] = grouped['team'].first()
stats = stats.reset_index().set_index(['team', 'player_id']).round(2)
stats

Out[10]:

		duration	distance	dist_1min	zone1_dist	zone2_dist	zone3_dist	zone4_dist	zone5_dist	max_speed	mean_x	mean_y
team	player_id
Home	1	1868.24	3086.51	99.13	1217.93	1282.80	455.48	120.13	10.18	26.04	45.94	50.91
	2	5800.24	9774.33	101.11	3725.63	4825.40	927.75	212.00	83.55	32.98	38.05	41.72
	3	5800.24	9498.54	98.26	4081.89	4252.98	878.45	171.69	113.52	32.59	38.08	30.83
	4	5800.24	10184.80	105.36	3687.15	4680.77	1067.73	536.40	212.75	32.42	40.74	18.66
	5	5800.24	11006.80	113.86	3847.24	5015.49	1130.58	647.68	365.81	32.54	53.73	46.08
	6	4433.96	8661.77	117.21	2766.99	4328.73	1307.11	232.44	26.51	26.15	51.25	35.93
	7	5800.24	10543.08	109.06	3885.76	4575.69	1507.00	490.91	83.72	28.30	44.25	46.68
	8	5800.24	11119.62	115.03	3898.15	4464.58	1522.57	783.53	450.79	31.98	57.90	16.37
	9	5800.24	10066.20	104.13	3951.72	3569.71	1467.12	768.27	309.38	30.26	63.93	33.78
	10	4855.68	9080.47	112.20	3910.21	3267.43	1094.72	607.13	200.98	32.68	61.60	34.41
	11	5800.24	3696.16	38.23	3282.56	371.18	33.80	6.24	2.38	27.73	11.37	33.99
	12	3932.04	7320.10	111.70	2511.08	3625.27	980.23	192.13	11.39	26.32	48.02	33.81
	13	1366.32	2439.28	107.12	945.75	889.51	279.18	224.78	100.05	32.03	60.26	35.99
	14	944.60	1925.19	122.29	564.13	775.15	351.83	224.01	10.05	25.24	52.88	49.74
Away	15	5800.24	9770.81	101.07	4132.80	4206.38	935.26	336.53	159.83	30.99	66.68	24.15
	16	5800.24	9836.75	101.76	4024.26	4247.96	1016.14	391.78	156.62	30.39	67.90	35.27
	17	5800.24	10849.34	112.23	4147.27	4247.13	1774.46	546.59	133.89	33.12	61.57	28.83
	18	5800.24	11165.79	115.50	3526.73	5295.39	1564.46	535.28	243.93	33.85	51.65	26.32
	19	4775.24	10746.20	135.02	3677.54	4506.57	2028.47	453.38	80.25	28.41	53.91	30.44
	20	5800.24	10218.13	105.70	3395.84	4754.24	1340.28	567.47	160.30	29.35	63.26	41.33
	21	5800.24	10893.03	112.68	3499.44	4977.85	1769.06	501.45	145.23	29.54	55.78	38.46
	22	4274.76	9283.93	130.31	4505.33	3048.55	1047.32	521.78	160.95	29.16	59.80	53.61
	23	5800.24	10222.69	105.75	3963.60	4291.65	1110.69	440.80	415.95	34.19	41.80	37.31
	24	4263.96	8620.45	121.30	3280.35	3463.72	1309.65	445.63	121.10	30.92	43.60	34.88
	25	5800.24	4867.23	50.35	3852.08	890.74	108.89	15.52	0.00	23.16	92.38	35.22
	26	1536.32	2788.39	108.90	999.14	1233.88	374.67	167.18	13.53	26.29	38.14	33.89
	27	1525.52	2844.05	111.86	930.09	1222.83	472.29	173.72	45.11	28.16	49.09	51.68
	28	1025.04	2107.81	123.38	620.93	851.56	361.36	231.88	42.09	26.00	50.40	36.72

활동량 지표 시각화¶

(1) 선수별 지표값 막대그래프 시각화¶

In [11]:

# 비교하고자 하는 컬럼
col_name = 'distance'

plt.figure(figsize=(15, 8))
plt.rcParams.update({'font.size': 15})

for team, color in zip(['Home', 'Away'], ['r', 'b']):
    team_stats = stats.loc[team]
    plt.bar(team_stats.index, team_stats[col_name], color=color, label=team)

plt.grid(axis='y', color='k', linestyle='--')
plt.legend()
plt.xticks(stats.reset_index()['player_id'])
plt.xlabel('player_id')
plt.ylabel(col_name)
plt.show()

(2) 속도 구간별 뛴 거리 막대그래프 시각화¶

In [12]:

plt.get_cmap('jet')

Out[12]:

jet

under

bad

over

In [13]:

plt.figure(figsize=(15, 8))
plt.title('Distance by Speed Zone')

player_ids = stats.reset_index()['player_id']
n_zones = len(distances_by_speed.columns)
colors = plt.cm.jet(np.linspace(0.9, 0.1, n_zones))

bottom = 0
for i, zone_dist in enumerate(distances_by_speed.columns[::-1]):
    plt.bar(player_ids, stats[zone_dist], bottom=bottom, color=colors[i], label=f'Zone {5-i}')
    if i < n_zones - 1:
        bottom = bottom + stats[zone_dist]

plt.grid(axis='y', color='k', linestyle='--')
plt.axvline(14.5, color='k', linestyle='--')

plt.xticks(stats.reset_index()['player_id'])
plt.ylim(0, 12000)
plt.xlabel('player_id')
plt.ylabel('distance')

# plt.legend() : legend 순서 변경
handles, labels = plt.gca().get_legend_handles_labels()
plt.legend(handles[::-1], labels[::-1])

plt.show()

(3) 선수별 평균 위치 및 지표값 산점도 시각화¶

In [14]:

col_name = 'dist_1min'

values_sorted = stats[col_name].sort_values()
min_size = 600
max_size = 1200
min_values = values_sorted.iloc[8]
max_values = values_sorted.iloc[-1]
sizes = min_size + (stats[col_name] - min_values) / (max_values - min_values) * (max_size - min_size)

draw_pitch('white', 'black', size_x=15, size_y=10)

for team, color in zip(['Home', 'Away'], ['r', 'b']):
    team_stats = stats.loc[team]
    x = team_stats['mean_x']
    y = team_stats['mean_y']
    plt.scatter(x, y, c=color, s=sizes[team].clip(0))
    for p in team_stats.index:
        plt.text(x[p], y[p], p, color='w', ha='center', va='center', fontsize=15)

plt.show()

저작자표시 비영리 변경금지