def compare_IRR():
print "Loading model..."
try:
model = load_from_pickle('pickle/model.pkl')
except OSError, IOError:
print "Model not found. Initializing training process, this might take some time..."
model = initialize_model()
def compare_IRR():
print "Loading model..."
try:
model = load_from_pickle("pickle/model.pkl")
except OSError, IOError:
print "Model not found. Initializing training process, this might take some time..."
model = initialize_model()
def run_process():
'''
Main process for app, invoked when administrator is at /refresh. Collects,
processes and stores data, as well as calculates expected IRR and other
metrics for data table and charts.
Returns:
Loan details to populate data table at /index. numpy array.
'''
print "Requesting loan details..."
loan_results, loan_details = request_loan_data()
print "Inserting results of API request to database..."
try:
insert_into_mongodb(loan_results, loan_details)
except Exception:
print "MongoDB error, proceeding to next step..."
print "Loading model..."
try:
model = load_from_pickle('pickle/model.pkl')
except OSError, IOError:
print "Model not found. Initializing training process, this might take some time..."
model = initialize_model()
df_raw = pd.concat((df_3c, df_3b), axis=0)
print "Pre-processing data..."
df = process_features(df_raw)
print "Initializing model..."
model = StatusModel(model=RandomForestRegressor,
parameters={
'n_estimators': 100,
'max_depth': 10
})
print "Training model..."
try:
model = load_from_pickle('pickle/model.pkl')
except OSError, IOError:
print "Model not found. Training model, this might take some time..."
model.train_model(df)
dump_to_pickle(model, 'pickle/model.pkl')
print "Calculating IRR..."
int_rate_dict = {
'A1': 0.0603,
'A2': 0.0649,
'A3': 0.0699,
'A4': 0.0749,
'A5': 0.0819,
'B1': 0.0867,
'B2': 0.0949,
'B3': 0.1049,
df_3c = pd.read_csv("data/LoanStats3c_securev1.csv", header=True).iloc[:-2, :]
df_3b = pd.read_csv("data/LoanStats3b_securev1.csv", header=True).iloc[:-2, :]
except OSError, IOError:
print "Training data not found. Please install from https://www.lendingclub.com/info/download-data.action"
df_raw = pd.concat((df_3c, df_3b), axis=0)
print "Pre-processing data..."
df = process_features(df_raw)
print "Initializing model..."
model = StatusModel(model=RandomForestRegressor, parameters={"n_estimators": 100, "max_depth": 10})
print "Training model..."
try:
model = load_from_pickle("pickle/model.pkl")
except OSError, IOError:
print "Model not found. Training model, this might take some time..."
model.train_model(df)
dump_to_pickle(model, "pickle/model.pkl")
print "Calculating IRR..."
int_rate_dict = {
"A1": 0.0603,
"A2": 0.0649,
"A3": 0.0699,
"A4": 0.0749,
"A5": 0.0819,
"B1": 0.0867,
"B2": 0.0949,
"B3": 0.1049,
