class ObjectReader():
from DataReader import DataReader
#These are the file extentions that have more reading
#This is to
OptionalInts = [
"w3d",
"w3a",
"w3q"
]
variableTypes = []
def __init__(self, filename):
self.read = DataReader(filename)
self.variableTypes = [
self.read.int,
self.read.float,
self.read.float,
self.read.string
]
self.fileVersion = self.read.int()
self.originalInfo = self.readTable()
self.customInfo = self.readTable()
def readMod(self):
modInfo = {}
modInfo["ID"] = self.read.charArray(4)
varType = self.read.int()
if filename.split(".")[1] in self.OptionalInts:
modInfo["level"] = self.read.int()
modInfo["pointer"] = self.read.int()
print(varType)
modInfo["value"] = self.variableTypes[varType]()
self.read.int() #verify / end thing
return modInfo
def readObject(self):
objectData = {}
objectData["oldID"] = self.read.charArray(4)
objectData["newID"] = self.read.charArray(4)
modCount = self.read.int()
objectData["mods"] = []
for i in xrange(modCount):
objectData["mods"].append(self.readMod())
return objectData
def readTable(self):
tmpLen = self.read.int()
tmpInfo = []
if tmpLen > 0:
for i in xrange(tmpLen):
tmpInfo.append(self.readObject())
return tmpInfo
def __init__(self):
QtGui.QDialog.__init__(self)
self.setupUi(self)
self.datreader = DataReader()
self.Plotter = Plotter()
self.directory = os.getcwd()
self.WorkingD_label.setText(self.directory)
self.ShowFile_PB.clicked.connect(self.show_file_start) # shows first lines in the textbrowser
self.ReadSets_PB.clicked.connect(self.read_set) # reads all files that start with lineEdit and creates a dict in the Sets_Dict[set][file][column]
self.PlotFile_PB.clicked.connect(self.plotfile)
self.MAV_slider.valueChanged.connect(self.mav_valuechanged)
self.MAV_slider.sliderReleased.connect(self.mav_released)
self.LP_slider.sliderReleased.connect(self.lp)
self.LP_slider.valueChanged.connect(self.lp_valuechanged)
self.HP_slider.sliderReleased.connect(self.hp)
self.HP_slider.valueChanged.connect(self.hp_valuechanged)
#self.CutZeros.clicked.connect(self.cut_zeros_filedict)
self.PlotColumn_PB.clicked.connect(self.plotcolumn)
self.Clear_PB.clicked.connect(self.clear)
self.Export_PB.clicked.connect(self.export)
self.FFT_PB.clicked.connect(self.fft)
self.ReadLabBook.clicked.connect(self.readlabbook)
self.MAVEdit.returnPressed.connect(self.mav_released)
self.MVAREdit.returnPressed.connect(self.mvar)
self.MMMINEdit.returnPressed.connect(self.mmmin)
self.Corr_PB.clicked.connect(self.correlate)
self.Select_PB.clicked.connect(self.open_filedialog)
self.Pyro_PB.clicked.connect(self.read_pyro)
self.Log_PB.clicked.connect(self.log_scale)
self.Sets_Dict = dict() # contains [set1][file1][column1] - the data
self.Files_Dict = dict() # contains [filename 1]: 'set-filename'
self.Columns_Dict = dict() # contains[set-filename-column]: same
def get_answer_part2():
result = 0
char_step = 0;
for paren in DataReader.read_data(1)[0]:
char_step += 1
result += 1 if paren == '(' else -1
if result < 0:
return char_step
def get_answer_part1(self):
#2572
position = Position(0, 0)
houses = {self.get_dict_key_from_position(position): True}
for move in DataReader.read_data(day=3)[0]:
position = AdventDayThree.adjust_pos_for_move(position, move)
houses[self.get_dict_key_from_position(position)] = True
return len(houses.keys())
def answer_part_one():
total_size = 0;
for wrapping in DataReader.read_data(day=2):
dimensions = wrapping.replace('\n', '').split('x')
int_dimensions = [int(dimensions[0]), int(dimensions[1]), int(dimensions[2])]
int_dimensions = sorted(int_dimensions)
total_size += 3 * (int_dimensions[0] * int_dimensions[1])
total_size += 2 * (int_dimensions[1] * int_dimensions[2])
total_size += 2 * (int_dimensions[2] * int_dimensions[0])
return total_size
def __init__(self, filename):
self.read = DataReader(filename)
self.variableTypes = [
self.read.int,
self.read.float,
self.read.float,
self.read.string
]
self.fileVersion = self.read.int()
self.originalInfo = self.readTable()
self.customInfo = self.readTable()
def get_answer_part2(self):
#2631
santa = Position(0, 0)
robot = Position(0, 0)
houses = {self.get_dict_key_from_position(santa): True}
move_count = 0
for move in DataReader.read_data(day=3)[0]:
move_count += 1
if move_count % 2:
santa = self.adjust_pos_for_move(santa, move)
houses[self.get_dict_key_from_position(santa)] = True
else:
robot = self.adjust_pos_for_move(robot, move)
houses[self.get_dict_key_from_position(robot)] = True
return len(houses.keys())
class PerceptronModel:
class TrainSplit:
def __init__(self):
self.train = []
self.test = []
def __init__(self):
self.reader = DataReader()
self.list_of_entities = []
self.list_of_klasses = []
self.num_of_folds = 10
def readTitles(self):
self.list_of_entities = self.reader.get_list_of_entities()
def crossValidationSplits(self):
splits = []
for fold in range(0, self.num_of_folds):
split = self.TrainSplit()
length = len(self.list_of_entities)
for i in range(0, length):
if i % self.num_of_folds == fold:
split.train.append(self.list_of_entities[i])
else:
split.test.append(self.list_of_entities[i])
splits.append(split)
return splits
def classify(self, entity):
pass
def train(self, split):
pass
def test(self, split):
pass
def __init__(self):
self.DR = DataReader()
self.Features = RuleBasedFeatures()
self.feats = ["cc", "ck", "bk", "pk", "hk", "oe", "3", "5", "6", "x", 'nword', 'hood', 'bCaret', 'cCaret', 'pCaret', 'hCaret']
self.gang = {}
use_cuda = torch.cuda.is_available()
# use_cuda = False
data_path = "../ivd_data/preprocessed.h5"
indicies_path = "../ivd_data/indices.json"
images_features_path = "../ivd_data/image_features.h5"
crop_features_path = "../ivd_data/image_features_crops.h5"
ts = str(datetime.datetime.fromtimestamp(time()).strftime('%Y_%m_%d_%H_%M'))
output_file = "logs/deciderguesser_output" + ts + ".log"
loss_file = "logs/decider_guesser_loss" + ts + ".log"
hyperparameters_file = "logs/deciderguesser_hyperparameters" + ts + ".log"
dr = DataReader(data_path=data_path, indicies_path=indicies_path, images_features_path=images_features_path, crop_features_path = crop_features_path)
### Hyperparamters
my_sys = getpass.getuser() == 'nabi'
length = 11
logging = True if my_sys else False
save_models = True if my_sys else False
# OpenNMT Parameters
opt = argparse.Namespace()
opt.batch_size = 1
opt.beam_size = 5
opt.gpu = 0
opt.max_sent_length = 100
opt.replace_unk = True
opt.tgt = None
class ReadW3E():
def __init__(self, filename):
self.read = DataReader(filename)
self.mapInfo = self.ReadMap()
def ReadMap(self):
mapInfo = self.ReadHeader()
mapInfo["info"] = []
for i in xrange((mapInfo["width"])*(mapInfo["height"])):
mapInfo["info"].append(self.ReadTile())
return mapInfo
def ReadHeader(self):
data = {}
data["fileID"] = self.read.charArray(4)
data["formatVersion"] = self.read.int()
data["mainTileSet"] = self.read.char()
data["customTileSet"] = self.read.int() #actually is a boolean
data["groundTileSets"] = self.ReadTileset()
data["cliffTileSets"] = self.ReadTileset()
data["width"] = self.read.int()
data["height"] = self.read.int()
data["offsetX"] = self.read.float()
data["offsetY"] = self.read.float()
return data
def ReadTileset(self):
length = self.read.int()
info = []
for i in range(0,length):
info.append(self.read.charArray(4))
return info
def ReadTile(self):
tmpData = {}
tmpData["groundHeight"] = self.read.short()
tmpData["waterLevel"] = self.read.short() #bit 15 is used for boundary flag 1
tmpData["nibble1"] = self.read.byte()
tmpData["textureDetails"] = self.read.byte()
tmpData["nibble2"] = self.read.byte()
return tmpData
test_set_num = 25
train_set_num = 660
valid_set_num = 7
path = ['.\data\A', '.\data\B', '.\data\C']
x_train = np.zeros((1, sample_size, 2))
x_test = np.zeros((1, sample_size, 2))
x_valid = np.zeros((1, sample_size, 2))
y_train = np.zeros((1))
y_test = np.zeros((1))
y_valid = np.zeros((1))
data = [x_train, x_test, x_valid, y_train, y_test, y_valid]
data_reader = DataReader(data_augment_stride=data_augment_stride,
train_set_num=train_set_num)
for i in range(len(path)):
data_sub = data_reader.read_data(path[i])
for i in range(len(data)):
data[i] = np.append(data[i], data_sub[i], axis=0)
for i in range(len(data)):
data[i] = data[i][1:]
# train set A, test set B
x_train, x_test, x_valid, y_train, y_test, y_valid = different_load_set(
data, 'B', 'A')
# for t_SNE
y_label = y_test
y_train = to_categorical(y_train, num_classes=10)
y_test = to_categorical(y_test, num_classes=10)
def get_answer_part1():
result = 0
for paren in DataReader.read_data(1)[0]:
result += 1 if paren == '(' else -1
return result
@author: haotianteng
"""
from DataReader import DataReader
#############################super parameter
TRANNING_READS = 30000 #Total reads used
TRAIN_WEIGHT = 0.4 #Proportion of reads used to train
TEST_WEIGHT = 0.4 #Proportion of reads used to test
VALID_WEIGHT = 0.2 #Proportion of reads used to validate
#Structure
EVENT_LENGTH = 20 #Length of each sentence
HIDDEN_UNIT_NUM = 24 #Length of the hidden state of each hidden layer
CellLayer = 3 #Number of the hidden layers
#Training
STEP_RATE = 0.5
BATCH_SIZE = 20
EPOCH = 5000
#############################
###############################Read the data
data = DataReader(
TRAIN_WEIGHT,
TEST_WEIGHT,
VALID_WEIGHT,
EVENT_LENGTH,
TRANNING_READS,
event_total=10000,
file_list='/home/haotianteng/UQ/BINF7000/Nanopore/GN_003/event_pass.dat')
Created on Oct 25, 2016
@author: Iegor
'''
from GraphAnimator import GraphAnimator
from DataReader import DataReader
from GreedySolver import solve_tsp
import numpy as np
def make_dist_matrix(points):
x = []
y = []
for point in points:
x.append(point[0])
y.append(point[1])
"""Creates distance matrix for the given coordinate vectors"""
N = len(x)
xx = np.vstack((x, ) * N)
yy = np.vstack((y, ) * N)
return np.sqrt((xx - xx.T)**2 + (yy - yy.T)**2)
if __name__ == "__main__":
d = DataReader('../../../capitals.txt')
points = d.readPoints()
matrix = make_dist_matrix(points)
path = solve_tsp(matrix)
g = GraphAnimator(points=points, pathes=path)
g.beginAnimation()
import os
import numpy as np
from sklearn.linear_model import LogisticRegression
from DataReader import DataReader
data_reader = DataReader("./data/us_trial.text", "./data/us_trial.labels")
X, Y = data_reader.get_features()
len_X = len(X)
len_X = (int)(len_X / 10)
len_train = (int)(0.8*len_X)
print(len_train)
trainX, trainY = X[0:len_train], Y[0:len_train]
trainX, trainY = np.asarray(trainX), np.asarray(trainY)
testX, testY = X[len_train:len_X], Y[len_train:len_X]
testX, testY = np.asarray(testX), np.asarray(testY)
print(len(testX))
if __name__ == '__main__':
clf = LogisticRegression(random_state=0, solver='lbfgs', multi_class='multinomial')
clf.fit(trainX, trainY)
counter, len_testX = 0, len(testX)
for i in range(len_testX):
output = clf.predict(np.asarray([testX[i]]))
if output == testY[i]:
counter += 1
accuracy = counter / len_testX
accuracy = accuracy * 100
def gen_dataset(seed):
random_seed = seed
data_reader = DataReader()
return DataObject(data_reader, 1, random_seed=random_seed)
# csvwriter.writerow(row)
# left[int(row[0])] -= 1
#input("Done")
connection_type = Connection_JIT(0, 0, 0)
kv_ty = (numba.types.containers.UniTuple(numba.int64,
2), numba.typeof(connection_type))
master_connection_dict = numba.typed.Dict.empty(*kv_ty)
innov_list = numba.typed.List()
innov_list.append(1)
print("Begin test...")
pr = Predictor()
p = Population(784, 47, 1, master_connection_dict, innov_list, 0)
dr = DataReader()
mapping = dr.get_mapping()
images = dr.get_images(
112800, 28, 28) # 112800 images in data set, each image 28x28 pixels
mp = p.networks[0]
for q in range(200):
print("Adding connection " + str(q + 1), end='\r')
mp.add_connection()
#mp.randomize_all_bias()
print("\nStarting predictions...")
engines = [[mp, 0, 0]]
pr.make_predictions(engines, mapping, images, 1)
base = engines[0][2]
from DataReader import DataReader
from Preprocessor import Preprocessor
from Vectorizer import Vectorizer
from Classifier import Classifier
from DeepLearning import DeepLearner
from sklearn.model_selection import train_test_split as split
import numpy as np
sub_b=['UNT','TIN']
dr_tr = DataReader('./datasets/training-v1/offenseval-training-v1.tsv','B')
tr_data,tr_labels = dr_tr.get_labelled_data()
tr_data,tr_labels = dr_tr.upsample(tr_data,tr_labels,label=0)
tr_data,tr_labels = dr_tr.shuffle(tr_data,tr_labels,'random')
dr_tst = DataReader('./datasets/test-B/testset-taskb.tsv')
tst_data,tst_ids = dr_tst.get_test_data()
tr_data = tr_data[:500]
tr_labels = tr_labels[:500]
##### Naive Bayes - Lemmatize - tfidf
prp = Preprocessor('remove_stopwords')
tr_data_clean = prp.clean(tr_data)
tst_data_clean = prp.clean(tst_data)
vct = Vectorizer('tfidf')
tr_vectors = vct.vectorize(tr_data_clean)
tst_vectors = vct.vectorize(tst_data_clean)
clf = Classifier('M-NaiveBayes')
def extractor(self):
#array in which to store people list
people_list = []
#initialization of prev_person
prev_person = ""
#for each person
for i in range(0, len(self.data_list)):
#identification of current person
curr_person = self.data_list[i].split("_")[0]
#if a new person is found
if curr_person != prev_person:
#if a previuous person exists
if prev_person != "":
#function to calculate features about a person
fc = FeaturesCalculator(self.strides_info)
#identification of gender: 0 for male, 1 for female
if prev_person[0] == 'M':
features = fc.calculator(0)
elif prev_person[0] == 'F':
features = fc.calculator(1)
else:
features = fc.calculator(-1)
#function to store features about a person
fs = FeaturesStorage(prev_person, features)
fs.storage()
#addind prev_person to list
people_list.append(prev_person)
#to free array with strides data
self.strides_info = []
#updating prev_person
prev_person = curr_person
#list of strides about current person
strides_list = os.listdir(self.strides_dir + self.data_list[i])
for j in range(0, len(strides_list)):
#reading data of the selected stride
dr = DataReader(self.strides_dir + self.data_list[i] +
'/' + strides_list[j] + '/kalman/txt/')
data, time = dr.reader()
#adding stride data to strides_info array
self.strides_info.append([time, data])
else:
#list of strides about current person
strides_list = os.listdir(self.strides_dir + self.data_list[i])
for j in range(0, len(strides_list)):
#reading data of the selected stride
dr = DataReader(self.strides_dir + self.data_list[i] +
'/' + strides_list[j] + '/kalman/txt/')
data, time = dr.reader()
#adding stride data to strides_info array
self.strides_info.append([time, data])
#if the last folder is rerached up
if i == len(self.data_list) - 1:
#function to calculate features about a person
fc = FeaturesCalculator(self.strides_info)
#identification of gender: 0 for male, 1 for female
if prev_person[0] == 'M':
features = fc.calculator(0)
elif prev_person[0] == 'F':
features = fc.calculator(1)
else:
features = fc.calculator(-1)
#function to store features about a person
fs = FeaturesStorage(prev_person, features)
fs.storage()
#addind prev_person to list
people_list.append(prev_person)
return people_list
import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
from HyperParameters import HyperParameters
from DataReader import DataReader
from NeuralNet import NeuralNet
def draw(reader, net):
plt.plot(reader.XTrain, reader.YTrain)
plt.show()
if __name__ == "__main__":
reader = DataReader()
reader.ReadData()
reader.NormalizeX()
reader.NormalizeY()
hp = HyperParameters(13,
1,
eta=0.001,
max_epoch=2000,
batch_size=50,
eps=1e-5)
net = NeuralNet(hp)
net.train(reader, checkpoint=0.2)
print("W=", net.weight)
print("B=", net.bias)
# lazy way for the cache decorator
@memoize
def _wordBreak(s):
# print(s)
results = []
for word in dict:
if s == word:
results.append(word)
elif s.startswith(word):
# print('got', word)
for result in _wordBreak(s[len(word):]):
results.append(word + ' ' + result)
return results
return _wordBreak(s)
def weight(s):
weight = 0
for word in s:
weight += 10**len(word)
return weight
if __name__ == "__main__":
d = DataReader('../../../dict_en.txt')
tokens = d.readTokens()
wraps = Solution().wordBreak("thisisatest", tokens)
print(sorted(wraps, key=weight))
def trainModel(experiment, testRun, setTarg):
print("Training model ...")
datasetTrain = DataReader(experiment.data["path"])
datasetTrain.setDatasetClassic("train", experiment.data["feature"],
experiment.data["annotation"])
if setTarg == "MeanStd": datasetTrain.setTargetMeanStd()
if testRun: datasetTrain = keepOne(datasetTrain)
datasetDev = DataReader(experiment.data["path"])
datasetDev.setDatasetClassic("dev", experiment.data["feature"],
experiment.data["annotation"])
if setTarg == "MeanStd": datasetDev.setTargetMeanStd()
if testRun: datasetDev = keepOne(datasetDev)
if testRun: experiment.maxEpoch = 1
inp, tar = datasetDev[0]
experiment.inputDim = inp.shape[1]
experiment.outputDim = tar.shape[1]
# print("experiment.outputDim", tar.shape)
wrapper = getWrapper(experiment)
wrapper.trainModel(datasetTrain,
datasetDev,
batchSize=experiment.batchSize,
maxEpoch=experiment.maxEpoch,
loadBefore=True,
tolerance=experiment.tolerance,
minForTolerance=experiment.minForTolerance)
wrapper.saveLogToCSV()
def __init__(self, file_path):
self.model = models.load_model(file_path)
self.data_reader = DataReader()
self.nn_helper_funcs = NNHelperFunctions()
self.history = []
for item_count in column_counts:
p = item_count / total
probabilities.append(p)
# print("probs", probabilities)
column_entropy = 0
for p_i in probabilities:
column_entropy += entropy(p_i)
# print("column_entropy", column_entropy)
return column_entropy
#accessing training data and setting up data frame
train_data_url = 'https://raw.githubusercontent.com/jeniyat/CSE-5521-SP21/master/HW/HW1/Data/train.csv'
dr = DataReader(train_data_url)
dr.read()
df = pd.DataFrame(dr.data)
#CALCULATING ENTROPY OF PREDICTION COLUMN
#getting edible column's counts of values
edible_col = df[0].value_counts().to_list()
probabilities = []
#get total count for current column
total = 0
for column in edible_col:
total += column
def __init__(self):
self.reader = DataReader()
self.list_of_entities = []
self.list_of_klasses = []
self.num_of_folds = 10
# tf.compat.v1.set_random_seed(seed_value)
# 5. Configure a new global `tensorflow` session
from keras import backend as K
#session_conf = tf.ConfigProto(intra_op_parallelism_threads=1, inter_op_parallelism_threads=1)
#sess = tf.Session(graph=tf.get_default_graph(), config=session_conf)
#K.set_session(sess)
session_conf = tf.compat.v1.ConfigProto(intra_op_parallelism_threads=1, inter_op_parallelism_threads=1)
sess = tf.compat.v1.Session(graph=tf.compat.v1.get_default_graph(), config=session_conf)
tf.compat.v1.keras.backend.set_session(sess)
from DataReader import DataReader
from NN_Manager import NN_Manager
#Get the dataSets
#tf.random.set_seed(13)
dataReader = DataReader()
(trainingDataSet,testingDataSet) = dataReader.GetDataSets()
(trainingLabels,testingLabels) = dataReader.GetLabels()
#Create chosen networks
networkManager = NN_Manager(trainingDataSet,testingDataSet,trainingLabels,testingLabels)
#networkManager.addNetwork("CNN")
#networkManager.addNetwork("LargerFilterCNN")
#networkManager.addNetwork("LargeCNN")
networkManager.addNetwork("StackedLSTM")
#networkManager.addNetwork("LSTM")
#networkManager.addNetwork("FFNN")
#networkManager.addNetwork("GRU")
#networkManager.addNetwork("StackedLSTM")
from DataReader import DataReader
import numpy as np
import pickle
import os
from time import time
data_path = '../ivd_data/preprocessed.h5'
indicies_path = '../ivd_data/indices.json'
dr = DataReader(data_path=data_path, indicies_path=indicies_path)
# source and target files
src_train = 'data/gw_src_train'
src_valid = 'data/gw_src_valid'
tgt_train = 'data/gw_tgt_train'
tgt_valid = 'data/gw_tgt_valid'
# parameters
length = 15
train_val_ratio = 0.1
n_games_to_train = '_ALL'
sos_token_stirng = '-SOS-'
# get all games
game_ids = dr.get_game_ids()
# prune games
def get_game_ids_with_max_length(length):
""" return all game ids where all questions are smaller then the given length """
# initial_population_size: number of chromosomes in the initial population
# nb_generations: nb of times we will produce a generation (includes tournament, cross over and mutation )
# ratio_cross : ratio of the total population to be crossed over and mutated
# prob_mutate : mutation probability
# k: number of participants on the selection tournaments.
initial_population_size, nb_generations, ratio_cross, prob_mutate, k = 200, 100, 0.8, 0.05, 2
# How many time to run the whole algo: for mini and small: 10 is enough, for big :100
problem_instances = 10
DATA_PATH = "data/data_transportationPb_mini.xlsx"
if __name__ == "__main__":
# reading the data from the excel file
dataReader = DataReader(DATA_PATH)
data = dataReader.read_data()
data["nb_vehicles"] = 9
# creating the list of vehicle names
vehicles = ['vehicle' + str(i) for i in range(0, int(data["nb_vehicles"]))]
stations = list(data["all_stations"])
distances = data["distances"]
mandatory_trips = data["trips"]
best_results = []
print("EXECUTING ", problem_instances, " INSTANCES ")
genetic_problem = GeneticProblem(vehicles, stations, distances, mandatory_trips)
generation_factory = GenerationFactory(genetic_problem)
t0 = time()
return model
def save_model(model_save_dir, model_name, model):
save_path = '{}/{}'.format(model_save_dir, model_name)
model.save(save_path)
if __name__ == '__main__':
source_path = 'data'
model_save_dir = 'models/'
img_width, img_height = 200, 200
model = create_vgg16_model(img_width, img_height)
data_reader = DataReader(source_path)
train_images, train_labels = data_reader.get_train_data()
val_images, val_labels = data_reader.get_val_data()
test_images, test_labels = data_reader.get_test_data()
np.set_printoptions(suppress=True)
history = model.fit(train_images,
train_labels,
validation_data=(val_images, val_labels),
batch_size=16,
epochs=10,
verbose=1)
predictions = model.predict(test_images)
for i in range(len(predictions)):
from DataReader import DataReader
x = DataReader('sign-ins.csv')
x.createPlot(x.createDatesList(), x.createCountList())
y = DataReader('sign-ins2.csv')
y.createPlot(y.createDatesList(), y.createCountList())
def read_header(self):
"""Read the header of a MPQ archive."""
def read_mpq_header(offset=None):
if offset:
self.file.seek(offset)
data = self.file.read(32)
header = MPQFileHeader._make(
struct.unpack(MPQFileHeader.struct_format, data))
header = header._asdict()
if header['format_version'] == 1:
data = self.file.read(12)
extended_header = MPQFileHeaderExt._make(
struct.unpack(MPQFileHeaderExt.struct_format, data))
header.update(extended_header._asdict())
return header
def read_mpq_user_data_header():
data = self.file.read(16)
header = MPQUserDataHeader._make(
struct.unpack(MPQUserDataHeader.struct_format, data))
header = header._asdict()
header['content'] = self.file.read(header['user_data_header_size'])
return header
magic = self.file.read(4)
self.file.seek(0)
print(magic)
if magic == "HM3W":
datReader = DataReader(self.file)
header = {}
## should be HM3W
header["wc3map_magic"] = datReader.charArray(4)
## unknown
datReader.int()
header["wc3map_mapName"] = datReader.string()
"""
0x0001: 1=hide minimap in preview screens
0x0002: 1=modify ally priorities
0x0004: 1=melee map
0x0008: 1=playable map size was large and has never been reduced to medium
0x0010: 1=masked area are partially visible
0x0020: 1=fixed player setting for custom forces
0x0040: 1=use custom forces
0x0080: 1=use custom techtree
0x0100: 1=use custom abilities
0x0200: 1=use custom upgrades
0x0400: 1=map properties menu opened at least once since map creation
0x0800: 1=show water waves on cliff shores
0x1000: 1=show water waves on rolling shores
"""
header["wc3map_mapFlags"] = datReader.flags()
header["wc3map_maxPlayers"] = datReader.int()
self.file.read(512 - datReader.index)
print ("Now position:", self.file.tell())
else:
## If the magic isn't HM3W, we will skip the first 512 bytes of the
## file anyway
self.file.seek(512)
print(self.file.tell())
magic = self.file.read(4)
self.file.seek(512)
print( len(magic))
print(magic, hex(ord(magic[3])) )
if magic == b'MPQ\x1a':
header.update(read_mpq_header())
header['offset'] = 512
elif magic == b'MPQ\x1b':
user_data_header = read_mpq_user_data_header()
header.update(read_mpq_header(user_data_header['mpq_header_offset']))
header['offset'] = user_data_header['mpq_header_offset']
header['user_data_header'] = user_data_header
else:
raise ValueError("Invalid file header.")
return header
pred_directory = os.path.join(directory, 'Pred') # make directory for storing predictions if it does not exist if os.path.isdir(pred_directory) == False: os.makedirs(pred_directory) # choose network, can be either DRN18 or DRN26 network = 'DRN26' # set parameters batch_size=8 num_epochs=100 use_weights = 1 num_classes = 5 image_dims=[500,500,3] data = DataReader(directory, batch_size, num_epochs, use_weights=0) dataset = data.test_batch(data_file) num_images = data.num_images # get image filenames image_list = data.image_list # determine number of iterations based on number of images num_iterations = int(np.floor(num_images/batch_size)) # create iterator allowing us to switch between datasets data_iterator = dataset.make_one_shot_iterator() next_element = data_iterator.get_next() # create placeholder for train or test train_network = tf.placeholder(tf.bool, [])
class Printer:
def __init__(self):
self.DR = DataReader()
self.Features = RuleBasedFeatures()
self.feats = ["cc", "ck", "bk", "pk", "hk", "oe", "3", "5", "6", "x", 'nword', 'hood', 'bCaret', 'cCaret', 'pCaret', 'hCaret']
self.gang = {}
def loadData(self, postsFile):
self.DR.loadData(postsFile)
def calculateFeatures(self, posts):
Hits = dd(int)
Scopes = dd(int)
numWordsScope = dd(int)
for post in posts:
numWords = len(self.DR.posts[post][4].split())
postHits, postScopes = self.Features.scorePostWordIndexing(self.DR.posts[post][4])
#print postHits, postScopes
for feat in self.feats:
Scopes[feat] += len(postScopes[feat])
Hits[feat + 'Count'] += len(postHits[feat + 'Count'])
numWordsScope[feat] += numWords
#print Hits, Scopes
#simpleGlobal = self.globalScoreSimple(Hits, Scopes)
#complexityGloabal = self.globalScoreComplexity(Hits, Scopes)
return Hits, Scopes, numWordsScope
def globalScoreComplexity(self, counts, scopeDict):
scopeIndices = set()
for feat in scopeDict.iterkeys():
for index in scopeDict[feat]:
scopeIndices.add(index)
count = 0
for feat in counts.iterkeys():
count += len(counts[feat])
if len(scopeIndices) > 0:
return str(round(count * 100.0 / len(scopeIndices), 2))
return ""
def globalScoreSimple(self, counts, scopeDict):
scope = 0
for feat in scopeDict.iterkeys():
scope += len(scopeDict[feat])
count = 0
for feat in counts.iterkeys():
count += len(counts[feat])
if scope > 0:
return str(round(count * 100.0 / scope, 2))
return ""
def printFeats(self, users, outFile):
outFile = open(outFile, 'w', 1)
for user in users:
Hits, Scopes, numWordsScope = self.calculateFeatures(self.DR.userwisePosts[user])
feats = []
for feat in self.feats:
try:
feats.append(str(round(Hits[feat+'Count'] * 100.0 / Scopes[feat], 2)))
except ZeroDivisionError:
feats.append('-1')
outFile.write(user + ',' + ','.join(feats) + ',' + self.gang[user] + '\n')
outFile.close()
def loadGangAnnotation(self, gangAnnotation):
for line in open(gangAnnotation):
line = line.strip().split('\t')
self.gang[line[0]] = line[1]
def _load_data(self):
self.raw = DataReader(day=7).as_raw()
def Train():
with tf.Graph().as_default(), tf.device('/cpu:0'):
global_step = tf.train.get_or_create_global_step()
inc_step = tf.assign_add(global_step, 1)
reader = DataReader(FLAGS.input_training_data,
FLAGS.buffer_size,
FLAGS.batch_size,
FLAGS.traing_epochs,
is_shuffle=True)
model = LRModel()
trainer = Trainer(model, inc_step, reader)
summary_op = tf.summary.merge_all()
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
config.allow_soft_placement = True
saver = tf.train.Saver(max_to_keep=FLAGS.max_model_to_keep,
name='model_saver')
with tf.Session(config=config) as session:
summ_writer = tf.summary.FileWriter(FLAGS.log_dir, session.graph)
#Load Pretrain
session.run(tf.local_variables_initializer())
session.run(tf.global_variables_initializer())
session.run(tf.tables_initializer())
session.run(reader.iterator.initializer)
#zero = tf.constant(0, dtype=tf.float32)
ckpt = tf.train.get_checkpoint_state(FLAGS.output_model_path)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(session, ckpt.model_checkpoint_path)
print("Load model from ", ckpt.model_checkpoint_path)
else:
print("No initial model found.")
trainer.start_time = time.time()
while True:
try:
_, avg_loss, total_weight, step, summary = session.run(
trainer.train_ops() + [summary_op])
#where = tf.not_equal(model.W, zero)
#indices = tf.where(where)
#print(session.run(indices))
#print(session.run(tf.gather(model.W, indices)))
#print(session.run(model.b))
#print(step)
if step % FLAGS.log_frequency == 1:
summ_writer.add_summary(summary, step)
trainer.print_log(total_weight, step, avg_loss)
if step % FLAGS.checkpoint_frequency == 1:
saver.save(session,
FLAGS.output_model_path + "/model",
global_step=step)
except tf.errors.OutOfRangeError:
print("End of training.")
break
pass
format(network_type, num_hidden_units, num_layers, num_epochs, embed_vector_size, window_size, min_count))
#mode = 'train'
mode = 'evaluate'
#mode = 'score'
K.clear_session()
with K.get_session() as sess:
K.set_session(sess)
graphr = K.get_session().graph
with graphr.as_default():
if mode == 'train':
print("Training the model... num_epochs = {}, num_layers = {}".
format(num_epochs, num_layers))
reader = DataReader(num_classes, vector_size=embed_vector_size)
entityExtractor = EntityExtractor(reader,
embedding_pickle_file)
entityExtractor.train(local_train_file_path,
network_type=network_type,
num_epochs=num_epochs,
num_hidden_units=num_hidden_units,
num_layers=num_layers)
entityExtractor.save_tag_map(tag_to_idx_map_file)
#Save the model
entityExtractor.save(model_file_path)
elif mode == 'evaluate':
# Evaluate the model
print("Evaluating the model...")
import scipy.sparse as sps
from Base.NonPersonalizedRecommender import TopPop, Random, GlobalEffects
from KNN.UserKNNCFRecommender import UserKNNCFRecommender
from KNN.ItemKNNCFRecommender import ItemKNNCFRecommender
from KNN.ItemKNNCBFRecommender import ItemKNNCBFRecommender
from GraphBased.RP3betaRecommender import RP3betaRecommender
from GraphBased.P3alphaRecommender import P3alphaRecommender
from Data_manager.RecSys2019.RecSys2019Reader import RecSys2019Reader
from Data_manager.DataSplitter_leave_k_out import DataSplitter_leave_k_out
if __name__ == '__main__':
for seed in [0, 1, 2, 3, 4]:
data_reader = DataReader()
data = DataObject(data_reader, 1, random_seed=seed)
recommender = RP3betaRecommender(data.urm_train)
recommender.fit(topK=10, alpha=0.27, beta=0.11)
LogToFileEvaluator.evaluate(data,
seed,
recommender,
"RP3",
"",
filename="algo_eval.csv")
# for seed in [0, 1, 2, 3, 4]:
# data_reader = DataReader()
# data = DataObject(data_reader, 1, random_seed=seed)
# recommender = ItemKNNCFRecommender(data.urm_train)
# recommender.fit(topK=22, shrink=850, similarity="jaccard", feature_weighting="BM25")
class MeinDialog(QtGui.QDialog, Dlg):
def __init__(self):
QtGui.QDialog.__init__(self)
self.setupUi(self)
self.datreader = DataReader()
self.Plotter = Plotter()
self.directory = os.getcwd()
self.WorkingD_label.setText(self.directory)
self.ShowFile_PB.clicked.connect(self.show_file_start) # shows first lines in the textbrowser
self.ReadSets_PB.clicked.connect(self.read_set) # reads all files that start with lineEdit and creates a dict in the Sets_Dict[set][file][column]
self.PlotFile_PB.clicked.connect(self.plotfile)
self.MAV_slider.valueChanged.connect(self.mav_valuechanged)
self.MAV_slider.sliderReleased.connect(self.mav_released)
self.LP_slider.sliderReleased.connect(self.lp)
self.LP_slider.valueChanged.connect(self.lp_valuechanged)
self.HP_slider.sliderReleased.connect(self.hp)
self.HP_slider.valueChanged.connect(self.hp_valuechanged)
#self.CutZeros.clicked.connect(self.cut_zeros_filedict)
self.PlotColumn_PB.clicked.connect(self.plotcolumn)
self.Clear_PB.clicked.connect(self.clear)
self.Export_PB.clicked.connect(self.export)
self.FFT_PB.clicked.connect(self.fft)
self.ReadLabBook.clicked.connect(self.readlabbook)
self.MAVEdit.returnPressed.connect(self.mav_released)
self.MVAREdit.returnPressed.connect(self.mvar)
self.MMMINEdit.returnPressed.connect(self.mmmin)
self.Corr_PB.clicked.connect(self.correlate)
self.Select_PB.clicked.connect(self.open_filedialog)
self.Pyro_PB.clicked.connect(self.read_pyro)
self.Log_PB.clicked.connect(self.log_scale)
self.Sets_Dict = dict() # contains [set1][file1][column1] - the data
self.Files_Dict = dict() # contains [filename 1]: 'set-filename'
self.Columns_Dict = dict() # contains[set-filename-column]: same
def log_scale(self):
print 'log_scale'
f = self.Plotter.figure_list[int(self.CurrentFigureEdit.text())]
ax = f.axes[int(self.ax_scale_edit.text())]
if ax.get_yscale == 'linear':
ax.set_yscale('log')
print 'log'
else:
ax.set_yscale('linear')
plt.title('')
def read_pyro(self):
print 'read_pyro'
filelist = list()
filelist = [f for f in os.listdir(self.directory) if f.startswith(self.lineEdit.text())]
print filelist
filelist = [os.path.join(self.directory, f) for f in filelist]
cols_of_interest = [str(c).rstrip(' ').lstrip(' ') for c in self.ColsOfInterestEdit.text().split(',')]
print cols_of_interest
self.Sets_Dict[str(self.lineEdit.text())] = self.datreader.read_pyro(filelist, cols_of_interest)
#self.cut_zeros_filedict()
self.update_SetScroll()
self.update_Files_Dict()
self.update_FileScroll()
self.update_Columns_Dict()
self.update_ColumnScroll()
print self.Sets_Dict.keys()
def open_filedialog(self):
files = str(QtGui.QFileDialog.getOpenFileName(None, QtCore.QString('Select File'), QtCore.QString(os.getcwd()),QtCore.QString('*.txt')))
print files
self.lineEdit.setText(os.path.basename(files))
self.WorkingD_label.setText(os.path.dirname(files))
self.directory = os.path.dirname(files)
def correlate(self):
fnum = self.Plotter.plot_eval(self.Plotter.correlate, 0, int(self.CurrentFigureEdit.text()), self.InActiveFigure.isChecked(), self.SelectedRange.isChecked(), self.SubtractMean_PB.isChecked())
self.CurrentFigureEdit.setText(str(fnum))
def readlabbook(self):
#self.HistoryEdit.insertPlainText(self.HistoryEdit.text())
print 'read labbook'
lab_dict = self.datreader.get_labdict(str(self.lineEdit.text()))
filelist = lab_dict.keys()
print filelist
path = self.directory
filelist = [os.path.join(path, f) for f in filelist if f in os.listdir(path)]
print filelist
cols_of_interest = [str(c).rstrip(' ').lstrip(' ') for c in self.ColsOfInterestEdit.text().split(',')]
print cols_of_interest
self.Sets_Dict[str(self.lineEdit.text())] = self.datreader.read_files(filelist, cols_of_interest)
#self.cut_zeros_filedict()
lab_dict = self.datreader.get_labdict(str(self.lineEdit.text()))
for f in self.Sets_Dict[str(self.lineEdit.text())].keys():
for info in lab_dict[f].keys():
self.Sets_Dict[str(self.lineEdit.text())][f][info] = lab_dict[f][info]
self.update_SetScroll()
self.update_Files_Dict()
self.update_FileScroll()
self.update_Columns_Dict()
self.update_ColumnScroll()
print self.Sets_Dict.keys()
def mvar(self):
fnum = self.Plotter.plot_eval(self.Plotter.mvar, int(self.MVAREdit.text()), int(self.CurrentFigureEdit.text()), self.InActiveFigure.isChecked(), self.SelectedRange.isChecked(), self.SubtractMean_PB.isChecked())
self.CurrentFigureEdit.setText(str(fnum))
def mmmin(self):
fnum = self.Plotter.plot_eval(self.Plotter.mmmin, int(self.MMMINEdit.text()), int(self.CurrentFigureEdit.text()), self.InActiveFigure.isChecked(), self.SelectedRange.isChecked(), self.SubtractMean_PB.isChecked())
self.CurrentFigureEdit.setText(str(fnum))
def mav_released(self):
#if not self.InActiveFigure.isChecked():
# self.MAVEdit.setText(str(self.MAV_slider.value()))
fnum = self.Plotter.plot_eval(self.Plotter.mav, int(self.MAVEdit.text()), int(self.CurrentFigureEdit.text()), self.InActiveFigure.isChecked(), self.SelectedRange.isChecked(), self.SubtractMean_PB.isChecked())
self.CurrentFigureEdit.setText(str(fnum))
def fft(self):
print 'fft'
fnum = self.Plotter.plot_eval(self.Plotter.fft, 0, int(self.CurrentFigureEdit.text()), self.InActiveFigure.isChecked(), self.SelectedRange.isChecked(), self.SubtractMean_PB.isChecked())
self.CurrentFigureEdit.setText(str(fnum))
def export(self):
self.Plotter.export(int(self.CurrentFigureEdit.text()))
def clear(self):
self.Sets_Dict = dict()
self.update_SetScroll()
self.Files_Dict = dict()
self.update_FileScroll()
self.Columns_Dict = dict()
self.update_ColumnScroll()
def lp_valuechanged(self):
self.LPEdit.setText(str(self.LP_slider.value()))
def lp(self):
print 'mav'
self.MAVEdit.setText(str(self.LP_slider.value()))
def hp_valuechanged(self):
self.HPEdit.setText(str(self.HP_slider.value()))
def hp(self):
print 'mav'
self.MAVEdit.setText(str(self.HP_slider.value()))
def mav_valuechanged(self):
self.MAVEdit.setText(str(self.MAV_slider.value()))
if self.InActiveFigure.isChecked():
fnum = self.Plotter.plot_eval(self.Plotter.mav, int(self.MAVEdit.text()), int(self.CurrentFigureEdit.text()), self.InActiveFigure.isChecked(), self.SelectedRange.isChecked(), self.SubtractMean_PB.isChecked())
self.CurrentFigureEdit.setText(str(fnum))
def plotcolumn(self):
for col in self.ColumnScroll.selectedItems():
key = str(col.text()).split('::')
col_data = self.Sets_Dict[key[0]][key[1]][key[2]]
x_axis = self.Sets_Dict[key[0]][key[1]]['Zeit']
label = str(col.text()+ '')
self.Plotter.plot_column(x_axis, col_data, int(self.CurrentFigureEdit.text()), label)
# def cut_zeros_filedict(self):
# print 'cut_zeros_filedict'
# if self.CutZeros.isChecked() == True:
# print 'checked'
# for fd in self.Files_Dict.keys():
# self.Files_Dict[fd] = self.datreader.cutzeros_file_dict(self.Files_Dict[fd])
def plotfile(self):
for f in self.FileScroll.selectedItems():
key = str(f.text()).split('::')
print key
title = str(f.text())
self.Plotter.plot_file(self.Sets_Dict[key[0]][key[1]], [ str(c).rstrip(' ').lstrip(' ') for c in self.ColsOfInterestEdit.text().split(',')], int(self.CurrentFigureEdit.text()), title)
def read_set(self):
print 'read_set'
filelist = list()
filelist = [f for f in os.listdir(self.directory) if f.startswith(self.lineEdit.text())]
print filelist
filelist = [os.path.join(self.directory, f) for f in filelist]
cols_of_interest = [str(c).rstrip(' ').lstrip(' ') for c in self.ColsOfInterestEdit.text().split(',')]
print cols_of_interest
self.Sets_Dict[str(self.lineEdit.text())] = self.datreader.read_files(filelist, cols_of_interest)
#self.cut_zeros_filedict()
self.update_SetScroll()
self.update_Files_Dict()
self.update_FileScroll()
self.update_Columns_Dict()
self.update_ColumnScroll()
print self.Sets_Dict.keys()
def update_ColumnScroll(self):
print 'update_ColumnScroll'
self.ColumnScroll.clear()
for col in self.Columns_Dict.keys():
item = QtGui.QListWidgetItem()
item.setText(col)
self.ColumnScroll.addItem(item)
def update_Columns_Dict(self):
print 'update_FilesDict'
cols_of_interest = [str(c).rstrip(' ').lstrip(' ') for c in self.ColsOfInterestEdit.text().split(',')]
for s in self.Sets_Dict.keys(): # sets
for f in self.Sets_Dict[s].keys(): #files
for c in self.Sets_Dict[s][f].keys():
if c in cols_of_interest:
self.Columns_Dict[s + '::' + f + '::' + c] = s + '::' + f + '::' + c
def update_Files_Dict(self):
print 'update_FilesDict'
for s in self.Sets_Dict.keys(): # sets
print s
for f in self.Sets_Dict[s].keys(): #files
print f
self.Files_Dict[f] = str(s) + '::'+ str(f)
#self.cut_zeros_filedict()
def update_SetScroll(self):
print 'update_SetScroll'
self.SetScroll.clear()
for key in self.Sets_Dict.keys():
item = QtGui.QListWidgetItem()
item.setText(str(key))
self.SetScroll.addItem(item)
def update_FileScroll(self):
print 'update_FileScroll'
self.FileScroll.clear()
for key in self.Files_Dict.keys():
item = QtGui.QListWidgetItem()
item.setText(str(self.Files_Dict[key]))
self.FileScroll.addItem(item)
def show_file_start(self):
try:
f = open(self.lineEdit.text())
s= ''
for i in range(12):
s = s+f.readline()
self.textBrowser.setText(s)
except:
print 'Error in file read'
def __init__(self, filename):
self.read = DataReader(filename)
self.mapInfo = self.ReadMap()
def get_tweets_labels(tweet_file, labels_file):
#Simply read in data
data_reader = DataReader(tweet_file, labels_file)
tweets = data_reader.read_tweets()
labels = data_reader.read_labels()
return tweets, labels
def main():
print("Running on BIO-NLP data\n\n")
from sys import platform
if platform == "win32":
home_dir = "C:\\dl4nlp"
else:
home_dir = os.path.join(os.path.expanduser('~'), "dl4nlp")
print("home_dir = {}".format(home_dir))
# The hyper-parameters of the word embedding trained model
window_size = 5
embed_vector_size = 50
min_count = 400
data_folder = os.path.join("sample_data", "drugs_and_diseases")
test_file_path = os.path.join(data_folder, "Drug_and_Disease_test.txt")
resources_pickle_file = os.path.join(home_dir, "models", "resources.pkl")
# The hyper-parameters of the LSTM trained model
#network_type= 'unidirectional'
network_type = 'bidirectional'
#embed_vector_size = 50
num_classes = 7 + 1
max_seq_length = 613
num_layers = 2
num_hidden_units = 150
num_epochs = 10
batch_size = 50
dropout = 0.2
reg_alpha = 0.0
print("Initializing data...")
model_file_path = os.path.join(home_dir,'models','lstm_{}_model_units_{}_lyrs_{}_epchs_{}_vs_{}_ws_{}_mc_{}.h5'.\
format(network_type, num_hidden_units, num_layers, num_epochs, embed_vector_size, window_size, min_count))
K.clear_session()
with K.get_session() as sess:
K.set_session(sess)
graphr = K.get_session().graph
with graphr.as_default():
# Evaluate the model
print("Evaluating the model...")
reader = DataReader(
input_resources_pickle_file=resources_pickle_file)
entityExtractor = EntityExtractor(reader)
#load the model
print("Loading the model from file {} ...".format(model_file_path))
entityExtractor.load(model_file_path)
entityExtractor.print_summary()
if not os.path.exists(os.path.join(home_dir, "output")):
os.makedirs(os.path.join(home_dir, "output"))
# make sure that the input test data file is in IOB format
output_prediction_file = os.path.join(home_dir, "output",
"prediction_output.tsv")
evaluation_report, confusion_matrix = entityExtractor.evaluate_model(
test_file_path, output_prediction_file)
print(evaluation_report)
print(confusion_matrix)
#########################################################
# from the commmand line interface,
# (1) change directory to \code\02_modeling\03_model_evaluation
# (2) run the following perl evaluation script
# "C:\Program Files\Git\usr\bin\perl.exe" Drug_and_Disease_eval.pl ..\..\..\sample_data\drugs_and_diseases\Drug_and_Disease_test.txt C:\dl4nlp\output\prediction_output.tsv
#########################################################
K.clear_session()
K.set_session(None)
print("Done.")
def populate_features(tweet_file, labels_file):
data_reader = DataReader(tweet_file, labels_file)
return data_reader.get_features()
from DataReader import DataReader
from Preprocessor import Preprocessor
from Vectorizer import Vectorizer
from Classifier import Classifier
from DeepLearning import DeepLearner
from sklearn.model_selection import train_test_split as split
import numpy as np
dr = DataReader('./datasets/training-v1/offenseval-training-v1.tsv', 'A')
data, labels = dr.get_labelled_data()
data, labels = dr.shuffle(data, labels, 'random')
data = data[:]
labels = labels[:]
prp = Preprocessor('remove_stopwords', 'lemmatize')
data = prp.clean(data)
tr_data, tst_data, tr_labels, tst_labels = split(np.array(data),
labels,
test_size=0.2,
stratify=labels)
tr_data, tr_labels = dr.upsample(tr_data, tr_labels, label=1)
tr_data, tr_labels = dr.shuffle(tr_data, tr_labels, 'random')
vct = Vectorizer('count')
vct.vectorize(tr_data)
model = DeepLearner(tr_data,
tr_labels,
vocab_length=vct.vocab_length,
class DataDriver:
def __init__(self, oscars):
self.Data = DataReader("tmdb-movies.csv")
self.Data.formatData()
self.OscarFile = pd.read_csv(oscars)
self.ActorsDictionary = {}
self.MovieDF = self.Data.getMovieDF()
self.Categories = [
"ACTOR", "ACTRESS", "ACTOR IN A SUPPORTING ROLE",
"ACTRESS IN A SUPPORTING ROLE", "ACTOR IN A LEADING ROLE",
"ACTRESS IN A LEADING ROLE"
]
self.OutputData = self.Data.getOutput()
self.cleanOscarData()
def scoreGenres(self):
genreList = [
'Action', 'Adventure', 'Science Fiction', 'Thriller', 'Fantasy',
'Crime', 'Western', 'Drama', 'Family', 'Animation', 'Comedy',
'Mystery', 'Romance', 'War', 'History', 'Music', 'Horror',
'Documentary', 'Foreign', 'TV Movie'
]
GenreScore = {
k: v
for (k,
v) in zip(genreList, list(reversed(range(len(genreList) +
1))))
}
for ind, row in self.MovieDF.iterrows():
score = 1
for genre in row["genres"]:
score *= GenreScore[genre]
self.Data.setNewAttribute(ind, "genres", score)
def setActorsDict(self):
for ind, row in self.MovieDF.iterrows():
for actor in row["cast"]:
self.ActorsDictionary[actor] = 0
self.scoreOscars()
def cleanOscarData(self):
self.OscarFile.drop(["year"], axis=1, inplace=True)
for ind, row in self.OscarFile.iterrows():
if row["category"] not in self.Categories:
self.OscarFile.drop([ind], inplace=True)
self.setActorsDict()
def scoreOscars(self):
for ind, row in self.OscarFile.iterrows():
if row["winner"]:
if row["entity"] in self.ActorsDictionary.keys():
val = self.ActorsDictionary[row["entity"]]
self.ActorsDictionary[row["entity"]] = val + 1
elif row["entity"] in self.ActorsDictionary.keys():
val = self.ActorsDictionary[row["entity"]]
self.ActorsDictionary[row["entity"]] = val
self.AddScores()
def SearchDict(self, dic, name):
for key, val in dic.items():
if key == name:
return val
return 0
def IterateScore(self, dic, arr):
Score = 0
for person in arr:
Score += self.SearchDict(dic, person)
return Score
def AddScores(self):
for ind, row in self.MovieDF.iterrows():
self.MovieDF = self.Data.setNewAttribute(
ind, "cast",
self.IterateScore(self.ActorsDictionary, row["cast"]))
def setRevOutput(self):
df1 = pd.get_dummies(self.MovieDF["revenue"])
self.MovieDF = pd.concat([self.MovieDF, df1], axis=1)
self.MovieDF.drop(["revenue"], axis=1, inplace=True)
def SaveData(self):
self.MovieDF.to_csv("DataRevExploration.csv", index=False)
import joblib
import numpy
from sklearn.preprocessing import MinMaxScaler
from DataReader import DataReader
if __name__ == '__main__':
# test_cases, test_result = DataReader.read_testdata()
test_cases, test_result = DataReader.read_testdata()
nn = joblib.load('./BPnn.pkl')
# 2021.01.07: 归一化与反归一化工作整合进BPmodel中
# 根据训练的模型预测测试集的结果
result_predict = nn.userPredict(test_cases)
i = 0
miss = 0
for res in result_predict:
print('No. ', i+1, res, test_result[i][0], '%.2f%%' %
(abs(res[0] - test_result[i][0]) / test_result[i][0] * 100))
if round(res[0]) != test_result[i][0]:
miss += 1
i = i + 1
print('correct rate: ', (i-miss)/i*100, '%')