Caribbean_Premier_League_Analysis.py

#!/usr/bin/env python
# coding: utf-8

# In[127]:


#import all relevant libraries
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from matplotlib import rcParams
import seaborn as sns
from collections import defaultdict
get_ipython().run_line_magic('matplotlib', 'inline')


# In[128]:


#read in csv file with data
df=pd.read_csv('cpl data/all_matches.csv')

#manually change start date to correct type as parse_dates=True didn't work. 
df['start_date'] = pd.to_datetime(df['start_date'])


# In[129]:


#Sort data (ball by ball) from first season to last season (in chronological order)
sorted_data = df.sort_values(by=['start_date', 'innings', 'ball'])
df = sorted_data


# In[130]:


#Get info on data: 


# In[131]:


df.head()


# In[132]:


df.tail()


# In[133]:


df.info()


# In[134]:


df.size


# In[135]:


df.shape


# In[136]:


df.dtypes


# In[137]:


#Add a column for overs:
df['overs'] = np.floor(df['ball']) + 1

#Add a column for keeping track of runs per ball
df['runs_per_ball'] = df['extras'] + df['runs_off_bat']

Clean data: alter all null values
# In[138]:


#Fix all columns with some non-null values: 

#wides, no balls, byes, legbyes, penalty : if Null => 0 else 1 
#wicket_type: if Null => No wicket 
#player_dismissed: if Null => No dismissal
df = df.fillna({'wides':0.0, 'noballs':0.0, 'byes':0.0, 'legbyes':0.0,
                'penalty':0.0, 'wicket_type':'No wicket', 'player_dismissed':'No dismissal'})


# In[139]:


#Drop columns with only null values:
df = df.dropna(axis=1)


# In[140]:


df.info()


# In[141]:


#Alter non-null values


# In[142]:


#Change venues to include country and also to remove duplicate values
#eg: Providence Stadium and Providence Stadium, Guyana
df = df.replace('Providence Stadium', 'Providence Stadium, Guyana')
df = df.replace("Queen's Park Oval, Port of Spain", "Queen's Park Oval, Port of Spain, Trinidad")
df = df.replace('Warner Park, Basseterre', 'Warner Park, Basseterre, St Kitts')
df = df.replace('Sabina Park, Kingston',  'Sabina Park, Kingston, Jamaica')
df = df.replace('Central Broward Regional Park Stadium Turf Ground', 
                'Central Broward Regional Park Stadium Turf Ground, USA')
df = df.replace('Daren Sammy National Cricket Stadium, Gros Islet',
                'Daren Sammy National Cricket Stadium, Gros Islet, St Lucia')
df = df.replace('Brian Lara Stadium, Tarouba', 'Brian Lara Stadium, Tarouba, Trinidad')
df = df.replace('Sir Vivian Richards Stadium, North Sound', 'Sir Vivian Richards Stadium, North Sound, Antigua')
df = df.replace("National Cricket Stadium, St George's", "National Cricket Stadium, St George's, Grenada")


# In[143]:


#Broad Analysis


# In[144]:


#List of Seasons: 
df['season'].unique()


# In[145]:


#List of Venues: 
df['venue'].unique()


# In[146]:


#List of ALL Teams (Not Franchises): 
df['batting_team'].unique()


# In[147]:


#Note - Franchise names were changed for some teams:

#Trinidad & Tobago Red Steel rebranded to Trinbago Knight Riders in 2015 so for the purpose of the analysis we will use Trinbago Knight Riders

#Barbados Tridents rebranded to Barbados Royals in July 2021 so for the purpose of the analysis we will use Barbados Royals

#St Lucia Zouks rebranded to St Lucia Stars in 2017 and later St Lucia Kings in 2021 so for the purpose of the analysis we will use St Lucia Kings. 


# In[148]:


#Change all Franchise names as indicated above:
df = df.replace('Trinidad & Tobago Red Steel', 'Trinbago Knight Riders')
df = df.replace("Barbados Tridents", "Barbados Royals")
df = df.replace('St Lucia Zouks', 'St Lucia Kings')
df = df.replace('St Lucia Stars', 'St Lucia Kings')


# In[149]:


#List of All Franchises: 
df['batting_team'].unique()


# In[150]:


# First ball of CPL history:
df.iloc[0]


# In[151]:


#Bar Chart showing Number of CPL matches played, in each season:

#Plot size
plt.figure(figsize=(12,10))

#Group by the match id and season columns
data = df.groupby(['match_id','season'])

#Count the data and drop the match_id 
data = data.count().index.droplevel(level=0).value_counts()

#Sort from 2013 - 2022
data = data.sort_index()

plt.title("Number of CPL matches per Season", bbox={'facecolor':'0.8', 'pad':5}, fontsize=12, y =1.0)

bar_chart = sns.barplot(y=data.index,x=data, orient='h')

bar_chart.bar_label(bar_chart.containers[0])

#Label Axes
plt.xlabel('Number of Matches Played', fontsize = 13)
plt.ylabel('Season', fontsize=14)
plt.rc('xtick', labelsize=10)
plt.rc('ytick', labelsize=10)

# displaying chart
plt.show()


# In[152]:


#Bar Chart showing Number of CPL matches played, at each Venue:

#Plot size
plt.figure(figsize=(12,12))

#Group by the match id and season columns
data = df.groupby(['match_id','venue'])

#Count the data and drop the match_id 
data = data.count().index.droplevel(level=0).value_counts()

#Sort from 2013 - 2022
data = data.sort_index()

plt.title("Number of CPL matches played at each Venue", bbox={'facecolor':'0.8', 'pad':5}, fontsize=16, y = 1.0)
rcParams['axes.titlepad'] = 50

bar_chart1 = sns.barplot(y=data.index,x=data)

#Label each bar 
bar_chart1.bar_label(bar_chart1.containers[0], fontsize=15)

#Label Axes
plt.xlabel('Number of Matches Played', fontsize=13)
plt.ylabel('Venue', fontsize=13)
plt.rc('xtick', labelsize=14)
plt.rc('ytick', labelsize=15)

# displaying chart
plt.show()


# In[153]:


#Bar Chart showing Number of CPL matches played, by each Franchise:

#Plot size
plt.figure(figsize=(12,10))

#Group by the match id and season columns
data = df.groupby(['match_id','batting_team'])

#Count the data and drop the match_id 
data = data.count().index.droplevel(level=0).value_counts()

#Sort from 2013 - 2022
data = data.sort_index()

plt.title("Number of CPL matches played by each franchise", bbox={'facecolor':'0.8', 'pad':7}, fontsize=14)

bar_chart2 = sns.barplot(y=data.index,x=data, orient='h')

bar_chart2.bar_label(bar_chart2.containers[0])

#Label Axes
plt.xlabel('Number of Matches Played', fontsize=12)
plt.ylabel('Franchise', fontsize=12)

plt.show()


# In[154]:


#Pie Chart showing Number of CPL matches played by each franchise:

#Group by the match id and team columns
data = df.groupby(['match_id','batting_team'])

#Count the data and drop the match_id 
data = data.count().index.droplevel(level=0).value_counts()

plt.title("Number of CPL matches played by each franchise", bbox={'facecolor':'0.8'}, y = 1.3, fontsize=12)

# define Seaborn color palette to use
palette_color = sns.color_palette('pastel')
  
# plotting data on chart
piechart2 = plt.pie(data, labels=data.index, colors=palette_color,         autopct=lambda p: '{:.1f}%({:.0f})'.format(p,(p/100)*data.sum()), radius = 1.75)
  
# displaying chart
plt.show()


# In[155]:


#CPL BATTING ANALYSIS


# In[156]:


#Total Runs Scored by each Franchise

#Get max runs for each team: sum of runs of the bat and the extras
team_runs = (df.groupby(['batting_team'])['runs_off_bat'].sum() + df.groupby(['batting_team'])['extras'].sum())/1000

#Sort the data in descending order; i.e. most runs first
team_runs = team_runs.sort_values(ascending=False)

plt.title("Total Runs scored by each Franchise (in Thousands)", bbox={'facecolor':'0.8', 'pad':7}, fontsize=14)

bar_chart3 = sns.barplot(y=team_runs.index,x=team_runs, orient='h')

bar_chart3.bar_label(bar_chart3.containers[0])

#Label Axes
plt.xlabel('Total Runs (Thousands)', fontsize=12)
plt.ylabel('Franchise', fontsize=12)

plt.show()


# In[157]:


#Top 10 Leading Run Scorers in the CPL
plt.figure(figsize=(10,8))

#Group by runs off the bat
batsman_runs = df.groupby(['striker'])['runs_off_bat'].sum()

#sort in descending order 
batsman_runs = batsman_runs.sort_values(ascending=False)[:10]


plt.title("Top 10 Leading Run Scorers in the CPL", bbox={'facecolor':'0.8', 'pad':6}, fontsize=14)

bar_chart4 = sns.barplot(y=batsman_runs.index,x=batsman_runs,orient='h')

bar_chart4.bar_label(bar_chart4.containers[0])

#Label Axes
plt.xlabel('Runs')
plt.ylabel('Batsman')
plt.show()


# In[158]:


#Produce Average Powerplay runs by each Franchise in a pivot Table

#Get powerplay overs
pp_overs = df[df['overs'] < 7]

#Get runs scored in pp overs
pp_runs = (pp_overs.groupby(['match_id', 'batting_team']).sum()['runs_off_bat'] + 
     pp_overs.groupby(['match_id', 'batting_team']).sum()['extras'])

##Group runs by franchise
pp_runs = pp_runs.groupby('batting_team')

#Get mean pp runs for each franchise and sort in ascending order
avg_pp_runs = pp_runs.mean().sort_values(ascending=False)

plt.title("Striplot showing Average Powerplay Score for each CPL Franchise", bbox={'facecolor':'0.8', 'pad':6}, fontsize=11)

#Create a Strip plot: Line graph where 1 data type is categorical and the other is numeric 
sns.stripplot(y=avg_pp_runs.index ,x=avg_pp_runs, hue=avg_pp_runs.index, data=avg_pp_runs)
print(avg_pp_runs)


# In[159]:


#Plot All CPL century- makers and number of centuries

plt.figure(figsize=(10,8))

#Get groups of batsman and amount of runs in each match
runs = df.groupby(['striker','match_id'])['runs_off_bat'].sum()
#Filter for centuries
centurians = runs[runs >= 100].droplevel(level=1).groupby('striker')

centurians = centurians.count().sort_values(ascending=False)

plt.title("Bar Chart showing all CPL Centurians", bbox={'facecolor':'0.8', 'pad':4}, fontsize=14)

bar_chart5 = sns.barplot(y=centurians.index,x=centurians,orient='h')

bar_chart5.bar_label(bar_chart5.containers[0])

#Label Axes
plt.xlabel('Number of Centuries', fontsize=12)
plt.ylabel('Batsman', fontsize=14)
plt.show()


# In[160]:


#Plot of the 10 Bastman with most CPL Half Centuries

plt.figure(figsize=(10,8))

#Get groups of batsman and amount of runs in each match
runs = df.groupby(['striker','match_id'])['runs_off_bat'].sum()

#Filter for 50s
half_centurians = runs[runs >= 50].droplevel(level=1).groupby('striker')

half_centurians = half_centurians.count().sort_values(ascending=False)[:10]

plt.title("Bar Chart of the 10 Bastman with the most CPL Half Centuries", bbox={'facecolor':'0.7', 'pad':4}, fontsize=14)

bar_chart6 = sns.barplot(y=half_centurians.index,x=half_centurians,orient='h')

bar_chart6.bar_label(bar_chart6.containers[0])

#Label Axes
plt.xlabel('Number of Half Centuries', fontsize=12)
plt.ylabel('Batsman', fontsize=14)
plt.show()


# In[161]:


#Batsman with the greatest number of runs each season

#Get runs for each player in every season
sum_data = df.groupby(['season','striker'])['runs_off_bat'].sum()

#Get seasons and max runs per season
season_max_runs = sum_data.groupby('season').max()

print("Batsman with the greatest number of runs each season:\n")
print("{0:10}{1:15}{2:1}".format("Season","Batsman","Runs"))

#traverse the season and runs dict (not really a dict)
#create a dataframe of sum of runs for each player per season
for season,run in season_max_runs.items():
    player = pd.DataFrame(sum_data).loc[season][pd.DataFrame(sum_data).loc[season]['runs_off_bat'] == run].index[0]
    print(season,'\t ',player,'\t', run)


# In[162]:


#List of the Top 10 greatest number runs scored by a single player in a season 

#Get runs for each player in every season
sum_data = df.groupby(['season','striker'])['runs_off_bat'].sum()

#Take top 10 most run scorers from any season
top_ten = sum_data.sort_values(ascending=False)[:10]

print("List of the Top 10 greatest number runs scored by a single player in a season\n")
print(top_ten)


# In[163]:


#10 Bastman with the most Fours struck
plt.figure(figsize=(10,8))

four_strikers = df[df['runs_off_bat'] == 4]['striker']

four_strikers = four_strikers.value_counts()[:10]

plt.title("Bar Chart of the 10 Bastman with the most Fours struck", bbox={'facecolor':'0.7', 'pad':4}, fontsize=14)

bar_chart7 = sns.barplot(y=four_strikers.index,x=four_strikers,orient='h')

bar_chart7.bar_label(bar_chart7.containers[0])

#Label Axes
plt.xlabel('Number of Fours', fontsize=12)
plt.ylabel('Batsman', fontsize=14)
plt.show()


# In[164]:


#Top 10 Highest Team Totals in CPL History
#Group data by start dates of matches and teams, i.e. 2 teams per match
team_runs = df.groupby(['start_date','batting_team']).sum()

#Get team totals for every game grouped by match dates
team_total = team_runs['runs_per_ball']

#Drop the match dates but keep the team and it's total score
team_total_scores = team_total.droplevel(level=0)

#Sort and get 10 highest
team_total_top = team_total_scores.sort_values(ascending=False)[:10]

plt.title("Striplot showing Top 10 Highest Team Totals in CPL History", bbox={'facecolor':'0.8', 'pad':6}, fontsize=11)

#Create a Strip plot: Line graph where 1 data type is categorical and the other is numeric 
stripplot2 = sns.stripplot(y=team_total_top.index ,x=team_total_top, hue=team_total_top.index,jitter=True, data=team_total_top)

plt.show()

#print pivot table
print(team_total_top)


# In[165]:


#10 Bastman with the most Sixes struck
plt.figure(figsize=(10,8))

six_strikers = df[df['runs_off_bat'] == 6]['striker']

six_strikers = six_strikers.value_counts()[:10]

plt.title("Bar Chart of the 10 Bastman with the most sixes struck", bbox={'facecolor':'0.7', 'pad':4}, fontsize=14)

bar_chart8 = sns.barplot(y=six_strikers.index,x=six_strikers,orient='h')

bar_chart8.bar_label(bar_chart8.containers[0])

#Label Axes
plt.xlabel('Number of Sixes', fontsize=12)
plt.ylabel('Batsman', fontsize=14)
plt.show()


# In[166]:


#Pie Chart showing Number of CPL Sixes Hit by Season

#Group by the match id and team columns
data = df[df['runs_off_bat'] == 6].groupby('season')

#Count the data
data = data.count()['match_id']

plt.title("Number of Sixes Hit in each CPL Season", bbox={'facecolor':'0.8'}, y = 1.3)

# define Seaborn color palette to use
palette_color = sns.color_palette('pastel')
  
# plotting data on chart
piechart1 = plt.pie(data, labels=data.index, colors=palette_color,       autopct=lambda p: '{:.1f}%({:.0f})'.format(p,(p/100)*data.sum()), radius = 1.75)

# displaying chart
plt.show()


# In[167]:


#Top 10 Highest individual Scores

#Get runs for each player in every innings in all seasons with dates
batter_date_score = df.groupby(['striker','start_date'])['runs_off_bat'].sum()

#Top 10 highest scores with batsman names and scores and also match dates
top_ten_scores_dates = batter_date_score.sort_values(ascending=False)[:10]

#Take top 10 highest scores with batsman names and scores only
top_ten_score = batter_date_score.sort_values(ascending=False)[:10].droplevel(level=1)

plt.title("Striplot showing the top 10 Highest individual Scores", bbox={'facecolor':'0.8', 'pad':6}, fontsize=13)

#Create a Strip plot: Line graph where 1 data type is categorical and the other is numeric 
stripplot4 = sns.stripplot(y=top_ten_score.index ,x=top_ten_score, hue=top_ten_score.index,jitter=True, data=top_ten_score)
plt.legend(bbox_to_anchor=(1, 1), loc=2)
plt.show()

#print pivot tables of scores, batsman names and dates
print(top_ten_scores_dates)



# In[168]:


#Top 10 individual innings with the Highest Number of 6s
#Get 6s for each player in every innings in all seasons with dates
six_data = df[df['runs_off_bat'] == 6].groupby(['start_date','striker'])

#Get num 6s hit by each player in every single match: match date: Player: #6s
six_count = six_data.count()['season']

most_sixes = six_count.sort_values(ascending=False).droplevel(level=0)[:10]

plt.title("Striplot showing the top 10 individual innings with the Highest Number of 6s", bbox={'facecolor':'0.8', 'pad':6}, fontsize=13)

#Create a Strip plot: Line graph where 1 data type is categorical and the other is numeric 
stripplot3 = sns.stripplot(y=most_sixes.index ,x=most_sixes.values, hue=most_sixes.index,jitter=True, data=most_sixes)
plt.legend(bbox_to_anchor=(1, 1), loc=2)
plt.show()

#print pivot table
print(most_sixes)


# In[ ]:





# In[169]:


#Bowling Data Analysis 


# In[170]:


#Leading wicket Takers by Season
#List of dismissals which aren't credited to bowler
no_cred_wickets = 'obstructing the field, retired hurt, No wicket, run out'

#Produce number of wickets for each bowler  in all seasons
bowler_wicket = df[df['wicket_type'].apply(lambda x: True if x not in no_cred_wickets else False)].groupby(['season','bowler'])
bowler_wicket_ct = bowler_wicket.count()['ball']

#sort for each season with highest number of wickets 1st
bowler_wicket_ct_seasonal=bowler_wicket_ct.sort_values(ascending=False)
best_bowler = bowler_wicket_ct_seasonal[:30].sort_index(level=0)
val=-1
lst=[]
max_wicket_taker_dict = {} 
max_wickets_dict = {}
print("Leading wicket Takers by Season with Number of Wickets")
print("{0:10}{1:20}{2:30}".format("Season","Player","Wickets"))

# Traverse best bowler: season:bowler: # wickets
#Keep highest wicket-taker
for (season,bowler),wicket in best_bowler.items():
    if season == val:
        #discard rest of entries below the highest 
        #for a season
        lst.append(wicket)
    else:
        #Create Printed Pivot Table
        print(season,'\t ',bowler,'\t\t',wicket)
        
        #season:best bowler
        max_wicket_taker_dict[season] = bowler
        
        #season:max wickets
        max_wickets_dict[season] = wicket
        val = season
        lst=[]

#Convert to Pandas Series

# #season:best bowler
s = pd.Series(max_wicket_taker_dict)

##season:max wickets
b = pd.Series(max_wickets_dict)


# In[171]:


#Bowlers with the most wickets each season (and number of wickets)
#Create a bar chart for max number of wickets each season
plt.title("Bar Chart of the most wickets each CPL season", bbox={'facecolor':'0.7', 'pad':4}, fontsize=14)

bar_chart9 = sns.barplot(y=b.index, x=b.values,orient='h')

bar_chart9.bar_label(bar_chart9.containers[0])

#Label Axes
plt.xlabel('Number of Wickets', fontsize=12)
plt.ylabel('Season', fontsize=14)
plt.show()


# In[172]:


#Create a Strip plot: Line graph where 1 data type is categorical and the other is numeric
#Striplot showing the Bowlers with most wickets each CPL season
plt.title("Striplot showing the Bowlers with most wickets each CPL season", bbox={'facecolor':'0.8', 'pad':6}, fontsize=13)

stripplot4 = sns.stripplot(y=s.values ,x=s.index, hue=s.values,jitter=True, data=s)
plt.legend(bbox_to_anchor=(1, 1), loc=2)
plt.show()


# In[173]:


#Top 10 Leading Wicket-Takers in the CPL
#List of dismissals which aren't credited to bowler
no_cred_wickets = 'obstructing the field, retired hurt, No wicket, run out'

#Get Data Set where if a bowler takes on a wicket on a ball; there is a True in that row
bw = df[df['wicket_type'].apply(lambda x: True if x not in no_cred_wickets else False)]

#produce series of bowlers and total wickets taken
#bw['bowler'] produces a series with bowlers names where the condition above evaluated True
#value_counts: counts all unique values
#Note: by default series is in descending order so no need to sort

bw = bw['bowler'].value_counts()[:10]

plt.figure(figsize=(10,8))

plt.title("Top 10 Leading Wicket-Takers in the CPL", bbox={'facecolor':'0.8', 'pad':6}, fontsize=14)

bar_chart11 = sns.barplot(y=bw.index,x=bw,orient='h')

bar_chart11.bar_label(bar_chart11.containers[0])

#Label Axes
plt.xlabel('Wickets')
plt.ylabel('Bowlers')
plt.show()


# In[174]:


#Number of Wickets Taken by each CPL Franchise

#List of dismissals which aren't credited to bowler
no_cred_wickets = 'obstructing the field, retired hurt, No wicket, run out'

#Get Data Set where if a bowler takes on a wicket on a ball; there is a True in that row
num_wickets = df[df['wicket_type'].apply(lambda x: True if x not in no_cred_wickets else False)]

#produce series of bowlers and total wickets taken
#bw['bowler'] produces a series with bowlers names where the condition above evaluated True
#value_counts: counts all unique values
#Note: by default series is in descending order so no need to sort

num_wickets = num_wickets['bowling_team'].value_counts()[:10]

plt.figure(figsize=(10,8))

plt.title("Number of Wickets Taken by each CPL Franchise", bbox={'facecolor':'0.8', 'pad':6}, fontsize=14)

bar_chart12 = sns.barplot(y=num_wickets.index,x=num_wickets,orient='h')

bar_chart12.bar_label(bar_chart12.containers[0])

#Label Axes
plt.xlabel('Wickets', fontsize=14)
plt.ylabel('Franchise', fontsize=14)
plt.show()


# In[175]:


#Bowlers with most dot balls

#Get Data Set of number of dots per bowler, grouped by bowler 
#All columms will have number of  dots bowled after running the count method so we can select from any at end
num_dots = df[df['runs_per_ball'] == 0].groupby('bowler').count()['runs_per_ball']

#Sort in descending order and pull top 10
num_dots = num_dots.sort_values(ascending=False)[:10]

plt.figure(figsize=(10,8))

plt.title("Bowlers with most dot balls bowled in CPL History", bbox={'facecolor':'0.8', 'pad':6}, fontsize=14)

bar_chart14 = sns.barplot(y=num_dots.index,x=num_dots,orient='h')

bar_chart14.bar_label(bar_chart14.containers[0])

#Label Axes
plt.xlabel('Number of Dot Balls', fontsize=14)
plt.ylabel('Bowler', fontsize=14)
plt.show()


# In[176]:


#Bowlers with most dot balls
#Striplot showing the Bowlers with most wickets each CPL season
plt.title("Striplot showing the Bowlers with most dot balls bowled in CPL History", bbox={'facecolor':'0.8', 'pad':6}, fontsize=13)

stripplot5 = sns.stripplot(y=num_dots.index ,x=num_dots.values, hue=num_dots.index,jitter=True, data=num_dots)
plt.legend(bbox_to_anchor=(1, 1), loc=2)
plt.show()


# In[177]:


#Bowlers with most CPL Maiden Overs
#Get dataframe  grouped by matches and bowlers showing the number of runs conceded
#in each over bowled by every bowler in that particular match.
#startdate:bowler:overs:runs in over
over_runs = df.groupby(['start_date','bowler','overs'])['runs_per_ball'].sum()

#filter all maiden overs:
over_runs = over_runs[over_runs.values == 0]

#Drop start date and overs: so datafeame will look like bowler:runs per over (always 0)
over_runs = over_runs.droplevel(level=[0,2])

#Now sum all maidens for each bowler
maidens = over_runs.index.value_counts()[:10]

plt.title("Striplot showing the Bowlers with most CPL Maiden Overs", bbox={'facecolor':'0.8', 'pad':6}, fontsize=13)

stripplot6 = sns.stripplot(y=maidens.index ,x=maidens.values, hue=maidens.index,jitter=True, data=maidens)
plt.legend(bbox_to_anchor=(1, 1), loc=2)
plt.show()


# In[178]:


#Number of Extras Conceded by All Franchises in CPL History

#Group Data Set by Bowling Team and sum extras given by Bowling Team. 
num_extras = df.groupby(['bowling_team'])['extras'].agg('sum')

#Sort in Descending Order
num_extras = num_extras.sort_values(ascending=False)

plt.figure(figsize=(10,8))

plt.title("Number of Extras Conceded by All Franchises in CPL History", bbox={'facecolor':'0.8', 'pad':6}, fontsize=14)

bar_chart15 = sns.barplot(y=num_extras.index,x=num_extras,orient='h')

bar_chart15.bar_label(bar_chart15.containers[0])

#Label Axes
plt.xlabel('Number of Extras', fontsize=14)
plt.ylabel('Franchise', fontsize=14)
plt.show()


# In[179]:


#Extras: Number of Wides Conceded by All Franchises in CPL History

#Group Data Set by Bowling Team and wides given by Bowling Team. 
num_wides = df.groupby(['bowling_team'])['wides'].agg('sum')

plt.title("Number of Wides bowled by each franchise in CPL History", bbox={'facecolor':'0.8'}, y = 1.3, fontsize=12)

# define Seaborn color palette to use
palette_color = sns.color_palette('pastel')
  
# plotting data on chart
piechart3 = plt.pie(num_wides, labels=num_wides.index, colors=palette_color,         autopct=lambda p: '{:.1f}%({:.0f})'.format(p,(p/100)*num_wides.sum()), radius = 1.75)
  
# displaying chart
plt.show()


# In[180]:


#Extras: No Balls Conceded by All Franchises in CPL History
num_noballs = df.groupby(['bowling_team'])['noballs'].agg('sum')

plt.title("Number of No Balls bowled by each franchise in CPL History", bbox={'facecolor':'0.8'}, y = 1.3, fontsize=12)

# define Seaborn color palette to use
palette_color = sns.color_palette('pastel')
  
# plotting data on chart
piechart4 = plt.pie(num_noballs, labels=num_noballs.index, colors=palette_color,         autopct=lambda p: '{:.1f}%({:.0f})'.format(p,(p/100)*num_wides.sum()), radius = 1.75)
  
# displaying chart
plt.show()


# In[181]:


#Top 10 Bowlers with Most No-Balls in CPL History

#Get All bowlers who bowled no balls
num_no_balls = df[df['noballs'] != 0]['bowler']

#Count Top 10 no ball bowlers
no_ball_top_10 = num_no_balls.value_counts()[:10]

plt.figure(figsize=(10,8))

plt.title("Top 10 Bowlers with Most No-Balls in CPL History", bbox={'facecolor':'0.8', 'pad':6}, fontsize=14)

bar_chart16 = sns.barplot(y=no_ball_top_10.index,x=no_ball_top_10 ,orient='h')

bar_chart16.bar_label(bar_chart16.containers[0])

#Label Axes
plt.xlabel('Number of No-Balls', fontsize=14)
plt.ylabel('Bowler', fontsize=14)
plt.show()


# In[182]:


#Bowlers with most runs conceded in an innings

#Get Data Set of bowler: match date: runs conceded in matc
runs_conceded_date = df.groupby(['bowler','start_date'])['runs_per_ball'].sum()

#Drop Match date:
runs_conceded = runs_conceded_date.droplevel(level=1)

#Sort in Descending Order
runs_conceded = runs_conceded.sort_values(ascending=False)[:10]

plt.figure(figsize=(10,8))

plt.title("Top 10 Bowlers with most runs conceded in an innings", bbox={'facecolor':'0.8', 'pad':6}, fontsize=14)

bar_chart17 = sns.barplot(y=runs_conceded.index,x=runs_conceded,orient='h')

bar_chart17.bar_label(bar_chart17.containers[0])

#Label Axes
plt.xlabel('Runs Conceded', fontsize=14)
plt.ylabel('Bowler', fontsize=14)
plt.show()

#Pivot Table
print(runs_conceded.sort_values(ascending=False)[:10])


# In[183]:


#Bowlers with most runs conceded in an innings
#Strip Plot 
plt.title("Striplot showing Top 10 Bowlers with most runs conceded in an innings", bbox={'facecolor':'0.8', 'pad':6}, fontsize=13)

stripplot8 = sns.stripplot(y=runs_conceded.index ,x=runs_conceded.values, hue=runs_conceded.index,jitter=True, data=runs_conceded)
plt.legend(bbox_to_anchor=(1, 1), loc=2)
plt.show()


# In[ ]:





# In[ ]:





# In[ ]:





# In[ ]:





# In[ ]: