6MeasuresOfFit.py

"""
Fitting the xG model

Need to run 3xGModel.py first to load data.
Then 5xGModelFit.py to fit the model.
"""

#Mcfaddens Rsquared for Logistic regression
null_model = smf.glm(formula="Goal ~ 1 ", data=shots_model, 
                           family=sm.families.Binomial()).fit()
1-test_model.llf/null_model.llf

#ROC curve
numobs=100
TP=np.zeros(numobs)
FP=np.zeros(numobs)
TN=np.zeros(numobs)
FN=np.zeros(numobs)


for i,threshold in enumerate(np.arange(0,1,1/numobs)):
    for j,shot in shots_model.iterrows():
        if (shot['Goal']==1): 
            if(shot['xG']>threshold):
                TP[i] = TP[i] + 1
            else:
                FN[i] = FN[i] + 1
        if (shot['Goal']==0): 
            if(shot['xG']>threshold):
                FP[i] = FP[i] + 1
            else:
                TN[i] = TN[i] + 1
     

fig,ax=plt.subplots(num=1)
ax.plot(FP/(FP+TN), TP/(TP+FN), color='black')
ax.plot([0,1], [0,1], linestyle='dotted', color='black')
ax.set_ylabel("Predicted to score and did TP/(TP+FN))")
ax.set_xlabel("Predicted to score but didn't FP/(FP+TN)")
plt.ylim((0.00,1.00))
plt.xlim((0.00,1.00))
ax.spines['top'].set_visible(False)
ax.spines['right'].set_visible(False)           
fig.savefig('Output/ROC_' + model  + '.pdf', dpi=None, bbox_inches="tight")