Peru World Cup Events Data Analysis¶

# Import modules

# Python ≥3.5 (ideally)
import platform
import sys
assert sys.version_info >= (3, 5)

# Import Dependencies
%matplotlib inline

# Math Operations
import numpy as np

# Data Preprocessing
import pandas as pd
import os    #  used to read the csv filenames
import re
import random

# Working with JSON
import json
from pandas.io.json import json_normalize

# Football libraries
## from FCPython import createPitch

# Data Visualisation
import matplotlib as mpl
import matplotlib.pyplot as plt
import seaborn as sns
## plt.style.use('seaborn-whitegrid')
from matplotlib.patches import Arc, Rectangle, ConnectionPatch
from matplotlib.offsetbox import OffsetImage
import missingno as msno    # not include with Conda, 'pip install missingno' in the terminal if you don't have it

import squarify    # pip install squarify
from functools import reduce

# Machine Learning
import scipy as sp

# Display in Jupyter
from IPython.display import Image, YouTubeVideo
from IPython.core.display import HTML

# Ignore warnings
import warnings
warnings.filterwarnings(action="ignore", message="^internal gelsd")

print("Setup Complete")

# Python / module versions used here for reference
print('Python: {}'.format(platform.python_version()))
print('NumPy: {}'.format(np.__version__))
print('pandas: {}'.format(pd.__version__))
print('matplotlib: {}'.format(mpl.__version__))
print('Seaborn: {}'.format(sns.__version__))
print('SciPy: {}'.format(sp.__version__))

competitions

# See the first match in the dataset - Peru vs. Australia
matches[0]

# See the away team for the first match in the dataset
matches[0]['away_team']

# See the away team name for the first match in the dataset
matches[0]['away_team']['away_team_name']

# Print all the match results
for match in matches:
    home_team_name = match['home_team']['home_team_name']
    away_team_name = match['away_team']['away_team_name']
    home_score = match['home_score']
    away_score = match['away_score']
    describe_text = f"The match between {home_team_name} and {away_team_name}"
    result_text = f" finished {home_score} : {away_score}"
    print(describe_text + result_text)

# Print match results involving Peru
for match in matches:
    home_team_name = match['home_team']['home_team_name']
    away_team_name = match['away_team']['away_team_name']
    if home_team_name == 'Peru' or away_team_name == 'Peru':
        home_score = match['home_score']
        away_score = match['away_score']
        describe_text = 'The match between ' + home_team_name + ' and ' + away_team_name
        result_text = ' finished ' + str(home_score) +  ' : ' + str(away_score)
        print(describe_text + result_text)

# Find ID for the match we are interested in - France vs. Peru
for match in matches:
    home_team_name = match['home_team']['home_team_name']
    away_team_name = match['away_team']['away_team_name']
    if (home_team_name == home_team_required) and (away_team_name == away_team_required):
        match_id_required = match['match_id']
print(home_team_required + ' vs ' + away_team_required + ' has id: ' + str(match_id_required))

fig=plt.figure()
fig.set_size_inches(7, 5)
ax=fig.add_subplot(1,1,1)
draw_pitch(ax)
plt.axis('off')
plt.show()

# France vs. Peru highlights, Match 21
YouTubeVideo('O4odLCih0Os')

# Get the nested structure into a DataFrame
df = pd.json_normalize(data, sep = "_").assign(match_id = file_name[:])

# DataFrame of Carrillo's actions on the pitch
carrillo_pass = df[(df['type_name'] == "Pass") & (df['player_name'] == 'André Martín Carrillo Díaz')]   # get passing information of Carrillo
carrillo_column = [i for i in df.columns if i.startswith("pass")]
carrillo_pass = carrillo_pass[["id", "period", "timestamp", "location", "pass_end_location", "pass_recipient_name"]]
carrillo_pass.head(60)

fig, ax = plt.subplots()
fig.set_size_inches(7, 5)
ax.set_xlim([0,120])
ax.set_ylim([0,80])
for i in range(len(carrillo_pass)):
    # can also differentiate by color
    color = "blue" if carrillo_pass.iloc[i]['period'] == 1 else "red"
    ax.annotate("", xy = (carrillo_pass.iloc[i]['pass_end_location'][0], carrillo_pass.iloc[i]['pass_end_location'][1]), xycoords = 'data',
               xytext = (carrillo_pass.iloc[i]['location'][0], carrillo_pass.iloc[i]['location'][1]), textcoords = 'data',
               arrowprops=dict(arrowstyle="->",connectionstyle="arc3", color = "blue"),)
plt.show()

# extract players involvement in the entire game
carrillo_action = df[(df['player_name'] == 'André Martín Carrillo Díaz')][["id", "type_name","period", "timestamp", "location"]]
carrillo_action.head()

fig, ax = plt.subplots()
fig.set_size_inches(7, 5)

x_coord = [i[0] for i in carrillo_action["location"]]
y_coord = [i[1] for i in carrillo_action["location"]]

#shades: give us the heat map we desire
# n_levels: draw more lines, the larger n, the more bluerry it loos
sns.kdeplot(x_coord, y_coord, shade = "True", color = "green", n_levels = 30)
plt.show()

heat_pass_map(df, 'André Martín Carrillo Díaz')

# we can see that ...

# ex. Ozil really struggles to play direct attacking ball in the first half, while he was a lot more direct in the second half

heat_pass_map(df, 'Christian Alberto Cueva Bravo')

heat_pass_map(df, 'José Paolo Guerrero González')

heat_pass_map(df, 'Luis Jan Piers Advíncula Castrillón')

shot_data = all_peru[(all_peru['type_name'] == "Shot") & (all_peru['team_name'] == 'Peru')]

fig = plt.figure()
fig.set_size_inches(7, 5)
ax = fig.add_subplot(1,1,1)
draw_pitch(ax)
plt.axis('off')

for i in range(len(shot_data)):
    # can also differentiate different half by different color
    color = "red" if shot_data.iloc[i]['shot_outcome_name'] == "Goal" else "black"
    ax.annotate("", xy = (shot_data.iloc[i]['shot_end_location'][0], shot_data.iloc[i]['shot_end_location'][1]), xycoords = 'data',
           xytext = (shot_data.iloc[i]['location'][0], shot_data.iloc[i]['location'][1]), textcoords = 'data',
           arrowprops=dict(arrowstyle = "->",connectionstyle="arc3", color = color),)
plt.ylim(0, 80)
plt.xlim(0, 120)
plt.show()

fig=plt.figure()
fig.set_size_inches(7, 5)
ax=fig.add_subplot(1,1,1)
draw_half_pitch(ax)
plt.axis('off')
# draw the scatter plot for goals
x_coord_goal = [location[0] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] == "Goal"]
y_coord_goal = [location[1] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] == "Goal"]
# shots that end up with no goal
x_coord = [location[0] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] != "Goal"]
y_coord = [location[1] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] != "Goal"]
# put the two scatter plots on to the pitch
ax.scatter(x_coord_goal, y_coord_goal, c = 'red', label = 'goal')
ax.scatter(x_coord, y_coord, c = 'blue', label = 'shots')
plt.ylim(0, 80)
plt.xlim(0, 120)
plt.legend(loc = 'upper right')
plt.axis('off')
plt.show()

# we use a joint plot to see the density of the shot distribution across the 2 axes of the pitch
joint_shot_chart = sns.jointplot(x_coord, y_coord, stat_func=None,
                                 kind = 'scatter', space=0, alpha=0.5)
joint_shot_chart.fig.set_size_inches(7,5)
ax = joint_shot_chart.ax_joint

# overlaying the plot with a pitch
draw_half_pitch(ax)
ax.set_xlim(0.5,120.5)
ax.set_ylim(0.5,80.5)

# draw the scatter plot for goals
x_coord_goal = [location[0] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] == "Goal"]
y_coord_goal = [location[1] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] == "Goal"]

# shots that end up with no goal
x_coord = [location[0] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] != "Goal"]
y_coord = [location[1] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] != "Goal"]

# put the two scatter plots on to the pitch
ax.scatter(x_coord, y_coord, c = 'b', label = 'shots')
ax.scatter(x_coord_goal, y_coord_goal, c = 'r', label = 'goal')

# Get rid of axis labels and tick marks
ax.set_xlabel('')
ax.set_ylabel('')
joint_shot_chart.ax_marg_x.set_axis_off()
ax.set_axis_off()
plt.ylim(-.5, 80)
plt.axis('off')
plt.show()

# I want to include some images into our diagram
peru = plt.imread("./img/farfanicon.png")
plt.imshow(peru)
plt.show()

cmap = plt.cm.YlOrRd_r    # import cmap
joint_shot_chart = sns.jointplot(x_coord, y_coord, stat_func=None,
                                 kind='reg', space=0, color = cmap(0.1))
joint_shot_chart.fig.set_size_inches(7,5)
ax = joint_shot_chart.ax_joint
draw_half_pitch(ax)
ax.set_xlim(0.5,120.5)
ax.set_ylim(0.5,80.5)

# draw the scatter plot for goals
x_coord_goal = [location[0] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] == "Goal"]
y_coord_goal = [location[1] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] == "Goal"]

# shots that end up with no goal
x_coord = [location[0] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] != "Goal"]
y_coord = [location[1] for i, location in enumerate(shot_data["location"]) if shot_data.iloc[i]['shot_outcome_name'] != "Goal"]

# put the two scatter plots on to the pitch
ax.scatter(x_coord, y_coord, c = 'b', label = 'shots')
ax.scatter(x_coord_goal, y_coord_goal, c = 'r', label = 'goal')
plt.legend(loc='lower right', bbox_to_anchor=(0.975, 0.0125))    # legend location specifically put here
plt.axis('off')

# Get rid of axis labels and tick marks
ax.set_xlabel('')
ax.set_ylabel('')
ax.set_title('Peru 2018 \nWorld Cup', 
             y=1.2, fontsize=15)
joint_shot_chart.ax_marg_x.set_axis_off()
joint_shot_chart.ax_marg_y.set_axis_off()
img = OffsetImage(peru, zoom=0.873)
img.set_offset((42,15.5)) # play around with the coordinate until I found a good place

# Add image of Farfan
ax.add_artist(img)
ax.set_axis_off()
plt.xlim(0,123)
plt.ylim(-.5, 83)
plt.axis('off')
plt.show()

# Export plot - not working currently
# plt.savefig('./img/fig/peru_all_shots_graphic.png')

data_id = [7562, 7546, 7532]
# consequently read the json and concatenate into a pre-defined dataframe

peru_all = pd.DataFrame()
for i in data_id:
    with open(dataDir + 'statsbomb/events/' + str(i)+'.json') as data_file:
        data = json.load(data_file)
    df = json_normalize(data, sep = '_')
    if all_peru.empty:
        peru_all = df
    else:
        peru_all = pd.concat([peru_all, df], join = 'outer', sort = False)

df_list = [total_dribble, dribble_complete, total_pass, pass_incomplete, dispossessed]
summary_data = reduce(lambda x, y: pd.merge(x, y, on = 'player_name'), df_list)
summary_data

# New dataframe, containing only players with more than 50 passes
dataPass= summary_data[summary_data["total_pass"]>50]

# Utilise matplotlib to scale our goal numbers between the min and max, then assign this scale to our values.
norm = mpl.colors.Normalize(vmin=min(dataPass.total_pass), vmax=max(dataPass.total_pass))
colors = [mpl.cm.Blues(norm(value)) for value in dataPass.total_pass]

# Create our plot and resize it.
fig = plt.gcf()
ax = fig.add_subplot()
fig.set_size_inches(16, 4.5)

# Use squarify to plot our data, label it and add colours. We add an alpha layer to ensure black labels show through
squarify.plot(label=dataPass.player_name,sizes=dataPass.total_pass, color = colors, alpha=.6)
plt.title("Passing data",fontsize=23,fontweight="bold")

# Remove our axes and display the plot
plt.axis('off')
plt.show()

dataDribble= summary_data[summary_data["total_dribble"]>0]

# Utilise matplotlib to scale our goal numbers between the min and max, then assign this scale to our values.
norm = mpl.colors.Normalize(vmin=min(dataDribble.total_dribble), vmax=max(dataDribble.total_dribble))
colors = [mpl.cm.Blues(norm(value)) for value in dataDribble.total_dribble]

# Create our plot and resize it.
fig = plt.gcf()
ax = fig.add_subplot()
fig.set_size_inches(16, 4.5)

# Use squarify to plot our data, label it and add colours. We add an alpha layer to ensure black labels show through
squarify.plot(label=dataDribble.player_name,sizes=dataDribble.total_dribble, color = colors, alpha=.6)
plt.title("Total dribble",fontsize=23,fontweight="bold")

# Remove our axes and display the plot
plt.axis('off')
plt.show()