fp_fc.py

from __future__ import absolute_import, division, print_function
import time
import warnings
import platform
import sys
import os

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
def warn(*args, **kwargs):
    pass
warnings.filterwarnings("ignore")
warnings.warn = warn

if platform.system() == 'Linux':
    sys.path.append("/home/apx748/OPV")
else:#It is Mac/Darwin
    sys.path.append("/Users/arindam/OPV")

import keras
from keras.models import Sequential
from keras.layers import Dense, Dropout
from keras.wrappers.scikit_learn import KerasRegressor
from ml_util import *


# Display training progress by printing a single dot for each completed epoch
class PrintEpochs(keras.callbacks.Callback):
    def on_epoch_end(self, epoch, logs):
        if epoch % 10 == 0:
            print('')
        else:
            print("epoch "+str(epoch)+" completed...")

def save_model(model, history, EPOCHS, lr, batch_size, fp="atom", data="cep"):
    file_suffix = "_"+fp+"_"+data+"_epochs_"+str(EPOCHS)+"_lr_"+str(lr)+"_batch_"+str(batch_size)
    model_file = join("model","model"+file_suffix+".json")
    weights_file = join("model","weights"+file_suffix+".h5")
    history_file = join("model","history"+file_suffix+".h5")

    model.save_weights(weights_file)
    # serialize model to JSON
    model_json = model.to_json()
    with open(model_file, "w") as json_file:
        json_file.write(model_json)
    saveNumpy(history,history_file)

    print("History saved successfully in", history_file)
    print("Model saved successfully in", model_file)
    print("Weights saved successfully in", weights_file)


def build_general_model(X_train, optimizer='adam', lr=0.001, dropout=0.0):
    '''
    It is used by morgan fingerprints as well as atom_v2 fingerprints
    '''
    model = Sequential()
    model.add(Dense(4096, activation= keras.activations.relu, input_dim = X_train.shape[1]))
    if dropout!=0.0:
        model.add(Dropout(dropout))
    model.add(Dense(1024, activation= keras.activations.relu))
    if dropout!=0.0:
        model.add(Dropout(dropout))
    model.add(Dense(256, activation= keras.activations.relu))
    if dropout!=0.0:
        model.add(Dropout(dropout))
    model.add(Dense(64, activation= keras.activations.relu))
    if dropout!=0.0:
        model.add(Dropout(dropout))
    model.add(Dense(1))

#     optimizer = tf.train.RMSPropOptimizer(0.001)

    if optimizer == "adam" and lr!= 0.001:
        print("Setting learning rate to"+str(lr))
        optimizer = tf.train.AdamOptimizer(lr)
    model.compile(loss='mse',
                optimizer=optimizer,
                metrics=['mape'])
    return model

def deep_regress_terminal(X,Y,EPOCHS=50,optimizer='adam',lr=0.001,dropout=False,deep=False,batch_size=32,fp="atom",data="cep"):
    X_train, X_test, y_train, y_test = train_test_split(X, Y, random_state=1024)
    if deep:
        print("We are building a deeper network")
    else:
        print("We are building default network")

    if len(X[0])==167: #if fingeprint is maccs or topo
        model = build_maccstopo_model(X_train, optimizer,lr,dropout,deep)
    elif len(X[0])==586: #if fingerprint is atom
        model = build_atom_model(X_train, optimizer,lr,dropout,deep)
    else:
        model = build_general_model(X_train, optimizer,lr,dropout)
    model.summary()

    # The patience parameter is the amount of epochs to check for improvement
    early_stop = keras.callbacks.EarlyStopping(monitor='val_loss', patience=5)
    # Store training stats
    start = time.time()
    if batch_size!=32:
        print("Batch size is set to: "+str(batch_size))
    history = model.fit(X_train, y_train, epochs=EPOCHS, batch_size=batch_size,
                        validation_split=0.2, verbose=0,
                        callbacks=[PrintEpochs()])
    end = time.time()
    print()
    print('Time elapsed:'+str(end-start)+' seconds')
    # plot_history(history)

    [loss, mape] = model.evaluate(X_test, y_test, verbose=0)

    print("Testing set Mean Abs Error: {:2.4f}".format(mape ))

    save_model(model, history, EPOCHS, lr, batch_size)

if __name__ == "__main__":
    ignored_compound_num = loadNumpy('missed_cep_ap')
    HOMO_orig = loadNumpy('HOMO_cep_all')
    HOMO = []
    for i in range(300000):
        if i in ignored_compound_num:
            continue
        HOMO += [HOMO_orig[i]]
    HOMO = np.array(HOMO)

    data="cep"
    fp = "atom"
    batch_size = 64
    if len(sys.argv)>2:
        batch_size = int(sys.argv[2])
        if len(sys.argv)>3:
            fp = sys.argv[3]
            if len(sys.argv)>4:
                data = sys.argv[4]

    if fp == "atom"
    atompair = loadNumpy('ap_cep_merged_1','atom_cep')
    for i in range(2,4):
        atompair = np.concatenate((atompair, loadNumpy('ap_cep_merged_'+str(i),'atom_cep')))
    print("Completed loading the atom pair fingerprint")
    atompair, HOMO = shuffle(atompair, HOMO, random_state=1024)
    print("Shuffling completed and training starting")



    model, history = \
    deep_regress_terminal(atompair, HOMO, int(sys.argv[1]), batch_size=64, fp=fp, data=data)
    print("Training complete")