def __init__(self, dataparser_obj, start, size):
self.dp = dataparser_obj
self.hyp = Hyperparameters()
self.labels = self.dp.get_meta()
self.size_of_sample = size
self.chunkIdentifier = str(size)
self.validation_start = 0
self.test_start = self.hyp.VALIDATION_NUMBER + self.size_of_sample - 1
self.train_start = self.test_start + self.hyp.TEST_NUMBER + self.size_of_sample - 1
self.train_matrix_data = list()
self.train_matrix_labels = list()
self.train_count = 0
self.valid_count = 0
self.test_count = 0
print("making validation set")
self.make_valid_set(start, size)
print("making test set")
self.make_test_set(start, size)
print("making train set")
self.make_train_set(start, size)
def __init__(self, dataparser_obj):
self.tool = Utility()
self.dp = dataparser_obj
self.hyp = Hyperparameters()
self.labels = self.dp.get_meta()
self.size_of_sample = self.dp.return_size_name(
self.hyp.MODE_OF_LEARNING)
self.chunkIdentifier = self.dp.return_chunkIdent_name(
self.hyp.MODE_OF_LEARNING)
self.validation_start = 0
self.test_start = self.hyp.VALIDATION_NUMBER + self.size_of_sample - 1
self.train_start = self.test_start + self.hyp.TEST_NUMBER + self.size_of_sample - 1
self.train_matrix_data = list()
self.train_matrix_labels = list()
self.train_count = 0
self.valid_count = 0
self.test_count = 0
print("making validation set")
self.make_valid_set()
print("making test set")
self.make_test_set()
print("making train set")
self.make_train_set()
def __init__(self, *argv):
self.hyp = Hyperparameters()
if(len((argv)) == 2):
print("I have registered a cookie-cutter chunk mode")
assert isinstance(argv[0], DataParser_Universal) == True, "you did not give me a DataParser object!"
assert isinstance(argv[1], str) == True, "you didn't give a proper key word"
self.chunkIdentifier = argv[1]
self.size = self.hyp.sizedict[self.chunkIdentifier]
self.mode = 1
elif(len(argv) == 3):
print("I have registered arbitrary mode")
assert isinstance(argv[0], DataParser_Universal) == True, "you did not give me a DataParser object!"
assert isinstance(argv[1], int) == True, "you didn't give a proper start"
assert isinstance(argv[2], int) == True, "you didn't give a proper size"
self.size = argv[2]
self.start = argv[1]
self.mode = 0
else:
raise Exception("invalid number of arguments")
self.dp = argv[0]
self.labels = self.dp.get_meta()
self.validation_start = 0
self.test_start = self.hyp.VALIDATION_NUMBER + self.size - 1
self.train_start = self.test_start + self.hyp.TEST_NUMBER + self.size - 1
self.train_matrix_data = list()
self.train_matrix_labels = list()
self.train_count = 0
self.valid_count = 0
self.test_count = 0
print("making validation set")
self.make_valid_set()
print("making test set")
self.make_test_set()
print("making train set")
self.make_train_set()
def __init__(self, dataparser_obj, arbiflag):
self.dp = dataparser_obj
self.hyp = Hyperparameters()
self.labels = self.dp.get_meta()
if not (arbiflag):
self.size_of_sample = self.dp.return_size_name(
self.hyp.MODE_OF_LEARNING)
self.chunkIdentifier = self.dp.return_chunkIdent_name(
self.hyp.MODE_OF_LEARNING)
self.validation_start = 0
self.test_start = self.hyp.VALIDATION_NUMBER + self.size_of_sample - 1
self.train_start = self.test_start + self.hyp.TEST_NUMBER + self.size_of_sample - 1
self.test_count = 0
print("making test set")
self.make_test_set()
def __init__(self):
self.hyp = Hyperparameters()
self.datasetList = list()
self.amparr = list()
self.superList = list(
) # this will contain dictionasries for each directory
for largeDirectory in self.hyp.data_to_include:
masteramparr = {}
files = sorted(listdir(largeDirectory))
for file in files:
if file.find(".") < 0:
try:
masteramparr[file] = self.getAmpArr(
file, largeDirectory)
print(largeDirectory + "/" + file)
self.datasetList.append(file)
except:
print(file + " was empty. I skipped it")
self.superList.append(masteramparr)
self.datasetList = list(dict.fromkeys(
self.datasetList)) # removes duplicates
import tensorflow as tf
import numpy as np
import csv
import os
from pipeline.ProjectUtility import Utility
import shutil
import pickle
from pipeline.MyCNNLibrary import * #this is my own "keras" extension onto tensorflow
from pipeline.Hyperparameters import Hyperparameters
from pipeline.DatasetMaker_Single_Test import DatasetMaker
from pipeline.DataParser_Single import DataParser
from housekeeping.csv_to_mat import ConfusionMatrixVisualizer
HYP = Hyperparameters()
DP = DataParser()
name = "Vanilla"
Cross = "test"
version = "AllDataCNN_test" + HYP.MODE_OF_LEARNING
weight_bias_list = list() #this is the weights and biases matrix
base_directory = "../Graphs_and_Results/" + name + "/" + version + "/"
try:
os.mkdir(base_directory)
print("made directory {}".format(
base_directory)) #this can only go one layer deep
except:
print("directory exists!")
pass