y['bio_rec_6_delay'] = data[['bio_rec_6_delay']].to_numpy().ravel()
nTest = 38
auc = np.zeros((len(tx), 3))
meanAuc = np.zeros((N, len(tx)))
cph = CoxPHSurvivalAnalysis()
sss = StratifiedShuffleSplit(n_splits=N, test_size=nTest)
for i, x in enumerate(tx):
x = x.to_numpy()
j = 0
for train, test in sss.split(x, y['bio_rec_6']):
xtrain, ytrain = x[train], y[train]
xtest, ytest = x[test], y[test]
cph.fit(xtrain, ytrain)
ypred = cph.predict(xtrain)
time = np.linspace(
np.min(ytrain['bio_rec_6_delay']) + 1,
np.max(ytrain['bio_rec_6_delay']) - 1, 3)
auc[i, :], meanAuc[j,
i] = cumulative_dynamic_auc(ytrain, ytrain, ypred,
time)
j += 1
meanAuc = pd.DataFrame(data=meanAuc,
columns=[
'4_feat.', "Prediction 1", 'Prediction 2',
'Prediction 3', 'Prediction 4'
])
meanMeanAuc = meanAuc.mean(axis=0)
stdMeanAuc = meanAuc.std(axis=0)
if model_name == 'NN':
model = MLPClassifier(hidden_layer_sizes=(100, 10))
if model_name == 'randomforest':
model = RandomForestClassifier(n_estimators=1000)
if model_name == 'gbm':
model = GradientBoostingClassifier()
if model_name == 'xgb':
model = XGBRegressor()
## Train and Predict
if (model_name == 'linearregression' or model_name == 'xgb'):
model.fit(X_train, Y_train)
Predict = model.predict(X_test)
elif (model_name == 'svmlin'):
model.fit(X_train, Y_train)
Predict = model.decision_function(X_test)
elif (model_name == 'coxregression'):
if data_name == 'maggic':
model.fit(Train_All,
duration_col='days_to_fu',
event_col='death_all')
Predict = model.predict_partial_hazard(X_test)
elif (data_name == 'heart_trans' or 'heart_wait'):
model.fit(Train_All,
duration_col="'Survival'",
event_col="'Censor'")