import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('https://jonghank.github.io/ase1302/files/FanGraphs_Leaderboard_2019.csv')
df

# your code here

data = np.array(df)
df['FIP'] = (13 * df['HR'] + 3*(df['BB']-df['IBB']+df['HBP']) -2*df['SO'])/df['IP'] + 3.20

plt.figure(figsize=(14,5), dpi=100)
plt.stem(df['Name'], df['FIP'], use_line_collection=True)
plt.xticks(rotation='vertical')
plt.ylabel('FIP')
plt.title('Fielding independent pitching')
plt.grid()
plt.show()


AL = ['Yankees', 'Rays', 'Red Sox', 'Blue Jays', 'Orioles', 'Twins', 'Indians', 'White Sox', 'Royals', 'Tigers', \
      'Astros', 'Athletics', 'Rangers', 'Angels', 'Mariners']
NL = ['Braves', 'Nationals', 'Mets', 'Phillies', 'Marlins', 'Cardinals', 'Brewers', 'Cubs', 'Reds', 'Pirates', 'Dodgers', \
      'Diamondbacks', 'Giants', 'Padres', 'Rockies']

# your code here

league = np.zeros(df.shape[0])
for i in range(df.shape[0]):
  if len(set(AL) & set([df['Team'][i]])):
    league[i] = 1
  elif len(set(NL) & set([df['Team'][i]])):
    league[i] = 2
  else:
    league[i] = 0
df['league'] = league
AL_index = np.where(df['league'] == 1)
NL_index = np.where(df['league'] == 2)
AL_name = df['Name'].values[AL_index]
NL_name = df['Name'].values[NL_index]
AL_top5 = AL_name[np.argsort(df['FIP'].values[AL_index])[:5]]
NL_top5 = NL_name[np.argsort(df['FIP'].values[NL_index])[:5]]

print(f'AL top 5 pitchers for FIP: {AL_top5}')
print(f'NL top 5 pitchers for FIP: {NL_top5}')

# your code here

CYP_TT = (df['IP']/2-df['ER']) + df['SO']/10 + df['W']
CYP_ESPN = 5*df['IP']/9 - df['ER'] + df['SO']/12 + 2.5*df['SV'] + df['ShO'] + 6*df['W'] - 2*df['L']
df['CYP_TT'] = CYP_TT
df['CYP_ESPN'] = CYP_ESPN 
AL_TT = AL_name[np.argmax(df['CYP_TT'].values[AL_index])]
NL_TT = NL_name[np.argmax(df['CYP_TT'].values[NL_index])]
AL_ESPN = AL_name[np.argmax(df['CYP_ESPN'].values[AL_index])]
NL_ESPN = NL_name[np.argmax(df['CYP_ESPN'].values[NL_index])]

print(f'By Tom Tango`s model, {AL_TT} in AL and {NL_TT} in NL will win the Cy Young Award')
print(f'By ESPN`s model, {AL_ESPN} in AL and {NL_ESPN} in NL will win the Cy Young Award')

plt.figure(figsize=(14,5), dpi=100)
plt.stem(df['Name'], df['CYP_TT'], use_line_collection=True)
plt.xticks(rotation='vertical')
plt.ylabel('Cy Young Award score (Tom Tango)')
plt.title('Cy Young Award prediction by Tom Tango')
plt.grid()
plt.show()

plt.figure(figsize=(14,5), dpi=100)
plt.stem(df['Name'], df['CYP_ESPN'], use_line_collection=True)
plt.xticks(rotation='vertical')
plt.ylabel('Cy Young Award score (ESPN)')
plt.title('Cy Young Award prediction by ESPN')
plt.grid()
plt.show()