In [1]:

    

%matplotlib inline


    

In [2]:

    

import cf2mongo as cf2m
import pandas as pd
import numpy as np
import seaborn as sns
from matplotlib import pyplot as plt
import seaborn as sns


    

In [3]:

    

crowdflower_data_path = "../Data/CrowdFlowerData/Football-Scenarios-DFE-832307.csv"
crowd_flower = cf2m.run_conversion(crowdflower_data_path)
crowd_flower.head()


    

    
Out[3]:







  

    

      

      
scenario
      
down
      
yards
      
field
      
quarter
      
clock
      
score
      
guess
    
  
  

    

      
0
      
It is first down and 10. The ball is on your o...
      
1.0
      
10.0
      
20.0
      
2.0
      
3.0
      
-3.0
      
fg
    
    

      
1
      
It is second down and inches. The ball is on y...
      
2.0
      
0.0
      
5.0
      
2.0
      
3.0
      
-3.0
      
fg
    
    

      
2
      
It is second down and inches. The ball is on y...
      
2.0
      
0.0
      
20.0
      
2.0
      
3.0
      
-3.0
      
fg
    
    

      
3
      
It is second down and inches. The ball is on y...
      
2.0
      
0.0
      
5.0
      
4.0
      
3.0
      
-3.0
      
fg
    
    

      
4
      
It is second down and inches. The ball is on y...
      
2.0
      
0.0
      
20.0
      
4.0
      
3.0
      
-3.0
      
fg
    
  

In [4]:

    

g = crowd_flower.groupby(['down','guess']).count().scenario.reset_index().pivot_table(index='down',columns='guess',values='scenario')
g['total'] = g.sum(axis=1)
g = g.reset_index()
g.columns.name = ''
down_guess = g.copy()
down_guess = down_guess[['down','pass','run','fg','punt','kneel','idk','total']]
down_guess


    

    
Out[4]:







  

    

      

      
down
      
pass
      
run
      
fg
      
punt
      
kneel
      
idk
      
total
    
  
  

    

      
0
      
1.0
      
148.0
      
49.0
      
10.0
      
NaN
      
2.0
      
8.0
      
217.0
    
    

      
1
      
2.0
      
477.0
      
521.0
      
63.0
      
NaN
      
10.0
      
13.0
      
1084.0
    
    

      
2
      
3.0
      
571.0
      
458.0
      
76.0
      
NaN
      
40.0
      
18.0
      
1163.0
    
    

      
3
      
4.0
      
131.0
      
115.0
      
456.0
      
408.0
      
33.0
      
21.0
      
1164.0
    
  

In [5]:

    

sns.violinplot(data=crowd_flower,x='quarter',y='guess')


    

    
Out[5]:





<matplotlib.axes._subplots.AxesSubplot at 0x2ea2347ac50>






    

In [6]:

    

sns.violinplot(data=crowd_flower,x='down',y='guess')


    

    
Out[6]:





<matplotlib.axes._subplots.AxesSubplot at 0x2ea238a91d0>






    

In [7]:

    

sns.violinplot(data=crowd_flower,x='field',y='guess')


    

    
Out[7]:





<matplotlib.axes._subplots.AxesSubplot at 0x2ea2373dda0>






    

In [8]:

    

def plot_PlayType(df,stat,playtypelist=['pass','run','fg','kneel','punt','idk'],percent_total=False):
    g = df.groupby([stat,'guess']).count().reset_index()
    g = g[g.columns[0:3]]
    last_col_name = g.columns[-1]
    g1 = g.groupby([stat, 'guess']).agg({last_col_name: 'sum'})
    if percent_total:
        g1 = g1.groupby(level=1).apply(lambda x: 100 * x / float(x.sum()))
    g1 = g1.reset_index()
    g1 = g1[g1.guess.apply(lambda x: x in playtypelist)]
    return sns.barplot(x=stat, y=last_col_name, hue="guess", data=g1)


    

In [9]:

    

fig = plt.figure(figsize=(16,32))
ax1 = fig.add_subplot(411)
ax1 = plot_PlayType(crowd_flower,'clock',['run','pass'],True)

ax2 = fig.add_subplot(412)
ax2 = plot_PlayType(crowd_flower,'clock',['run','pass'],True)

ax3 = fig.add_subplot(413)
ax3 = plot_PlayType(crowd_flower,'quarter',['run','pass'],True)

ax4 = fig.add_subplot(414)
ax4 = plot_PlayType(crowd_flower,'field',['run','pass'],True)