001_run_ML.py

from Utils import *
from Utils_parallel import *

SEED = 741
njob = 1

# exec(open("015_SPLITTING_DATA.py").read())
# exec(open("030_VARIABLES_SELECTION.py").read())
# exec(open("035_UNIVARIATE_VARIABLES_SELECTION.py").read())

nvars = np.arange(140, 252, 20) #np.concatenate( ([1], np.arange(10, 51, 10), np.arange(70, 140, 30)) )
methods = ['LR_ACCURACY', 'E_NET', 'INFORMATION_GAIN', 'LASSO', 'RIDGE', 'RANDOM_FOREST', 'GBM']


probs_to_check = np.arange(0.1, 0.91, 0.1)
DF = pd.DataFrame()

scheduled_model = 'running_model/'
create_dir( scheduled_model)

for method in methods:
    for nvar in nvars:
        predictors = extract_predictors(method, nvar, SEED)
        eff_nvar = len(predictors)
        print method, eff_nvar
        try:
            exec(open("041_TREE_BASED_MODELS.py").read())
            DF.to_csv(scheduled_model + 'OK_TREE_BASED_MODELS' + method + '_' + str(nvar) + '.csv')
        except:
            DF.to_csv( scheduled_model + 'ERROR_TREE_BASED_' + method + '_' + str(nvar) + '.csv')
        try:
            exec(open("043_REGULARIZED_METHODS.py").read())
            DF.to_csv(scheduled_model + 'OK_REGULARIZED_METHODS' + method + '_' + str(nvar) + '.csv')
        except:
            DF.to_csv( scheduled_model + 'ERROR_LASSO_' + method + '_' + str(nvar) + '.csv')
        try:
            exec(open("045_NAIVE_BAYES.py").read())
            DF.to_csv(scheduled_model + 'OK_NAIVE_BAYES' + method + '_' + str(nvar) + '.csv')
        except:
            DF.to_csv( scheduled_model + 'ERROR_NAIVE_BAYES_' + method + '_' + str(nvar) + '.csv')
        try:
            exec(open("046_KNN.py").read())
            DF.to_csv( scheduled_model + 'OK_KNN_' + method + '_' + str(nvar) + '.csv')
        except:
            DF.to_csv(  scheduled_model + 'ERROR_' + method + '_' + str(nvar) + '.csv' )
        try:
            exec (open("042_SVM.py").read())
            DF.to_csv(scheduled_model + 'OK_SVM' + method + '_' + str(nvar) + '.csv')
        except:
            DF.to_csv( scheduled_model + 'ERROR_SVM_' + method + '_' + str(nvar) + '.csv')


for method in methods:
    for nvar in nvars:
        predictors = extract_predictors(method, nvar, SEED)
        eff_nvar = len(predictors)
        print method, nvar
    try:
        exec(open("051_NEURAL_NETWORK.py").read())
        DF.to_csv(scheduled_model + 'OK_NN_' + method + '_' + str(nvar) + '.csv')
    except:
        DF.to_csv( scheduled_model + 'ERROR_NN_' + method + '_' + str(nvar) + '.csv')