main.py

#coding=utf-8

import os
import random
import numpy as np
import tensorflow as tf

from dataPro import *
from baseDNNModel import *
from modelTrainer import *
from generateEggs import *

flags = tf.app.flags

flags.DEFINE_string(
    "gpuId",
    "3",
    "Which gpu is assigned.")

flags.DEFINE_string(
    "fileRootPath",
    "./files",
    "File path for all files.")

flags.DEFINE_string(
    "dataRootPath",
    "./files",
    "Data file path for all data.")

flags.DEFINE_string(
    "path4SaveEggsFile",
    "./files",
    "The path for saving eggs file.")

flags.DEFINE_string(
    "path4Summaries",
    "./files/summaries",
    "The path for saving summaries.")

flags.DEFINE_string(
    "path4SaveModel",
    "./files/trainedModel",
    "The path for saving model.")

flags.DEFINE_string(
    "oneCLDataPath4Training",
    "./files/4_training/PC3",
    "The path for training in one cell line.")

flags.DEFINE_string(
    "oneCLDataPath4GenerateEggs",
    "./files/after_merge/PC3",
    "The path for generating eggs in one cell line.")

flags.DEFINE_float(
    "testSize",
     1e-1,
     "The threshold for validation data.")

flags.DEFINE_float(
     "learningRate",
     1e-4,
     "The learning rate for training.")

flags.DEFINE_float(
     "dropOutRate",
     0.5,
     "The threshold for validation data.")

flags.DEFINE_float(
     "threshold4Val",
     0.5,
     "The threshold for validation data.")

flags.DEFINE_float(
     "threshold4Convegence",
     1e-40,
     "The threshold for training convegence.")

flags.DEFINE_integer(
     "batchSize",
     150,
     "How many samples are trained in each iteration.")

flags.DEFINE_integer(
     "trainEpoches",
     1000,
     "How many times training through all train data.")

flags.DEFINE_integer(
     "nWeight",
     10,
     "The weighted for negative samples in objective funcion.")

FLAGS = flags.FLAGS

if __name__ == "__main__":
  os.environ["CUDA_VISIBLE_DEVICES"] = FLAGS.gpuId
  insDataPro = DataPro(FLAGS)
  insDataPro.loadDataInL1000()
  insDataPro.loadOneCLData()
  insDataPro.transferLabel2TwoCol()
  numOfNeurons = [1956, 200, 10, 2]
  insDNNModel = BaseDNNModel(FLAGS, numOfNeurons)
  insDNNModel.buildBaseDNNModelGraph()
  insModelTrainer = ModelTrainer(FLAGS, insDataPro, insDNNModel)
  modelSavePath = insModelTrainer.trainDNN()
  insGenerateEggs = GenerateEggs(FLAGS, insDataPro, modelSavePath)
  insGenerateEggs.generateEggs2Files()