import pandas as pd

import numpy as np

import utils as util

import matplotlib.pyplot as plt

if util.final:

columns, data = util.readFile('dating.csv')

else:

columns, data = util.readFile('test_dating.csv')

for col in util.rPP:

distinctValues = np.sort(data[col].unique())

success_rates={}

for val in distinctValues:

dfWithVal = data[data[col]==val]

dfWithValSuccess = dfWithVal[dfWithVal.decision==1]

success_rates[val] = len(dfWithValSuccess)/len(dfWithVal)

plt.scatter(distinctValues,

[success_rates[val] for val in distinctValues],

marker='o',

s=50)

plt.xlabel(col)

plt.ylabel('Success Rate')

if util.final:

plt.savefig('outputs/2_2/success_rate_'+col+'.pdf',format='pdf')

else:

plt.savefig('outputs/2_2/test_success_rate_'+col+'.pdf',format='pdf')

plt.clf()