def read_data():
##
logger = logging.getLogger(__name__)
logger.info('Reading data for from [{}]'.format(dt.data_dir(), 'task2'))
X_test = pandas.read_csv(os.path.join(dt.data_dir(), 'task2',
'X_test.csv'),
header=0,
index_col=0)
X_train = pandas.read_csv(os.path.join(dt.data_dir(), 'task2',
'X_train.csv'),
header=0,
index_col=0)
y_train = pandas.read_csv(os.path.join(dt.data_dir(), 'task2',
'y_train.csv'),
header=0,
index_col=0)
##
logger.info('Modify y_train to be of type <<int.>>')
y_train.y = y_train.y.astype(int)
return \
X_test,\
X_train,\
y_train,\
def read_data():
test = pandas.read_csv(os.path.join(dt.data_dir(), 'task0', 'test.csv'),
header=0,
index_col=0)
train = pandas.read_csv(os.path.join(dt.data_dir(), 'task0', 'train.csv'),
header=0,
index_col=0)
return test, train
def main():
## Read Data
trainingData = pandas.read_csv(os.path.join(cd.data_dir(),
cd.DataSets.EX1.value,
'train.csv'),
header=0,
index_col=0)
#rs = numpy.random.RandomState(99)
#newIndex = rs.choice(trainingData.index,trainingData.__len__())
#trainingData = trainingData.loc[newIndex, :]
yCols = ['y']
xCols = trainingData.drop(columns=yCols).columns
## Set-up cross validation
lambdaParam = [.1, 1, 10, 100, 1000]
rmseVec = pandas.Series(index=[str(i) for i in lambdaParam], name='RMSE')
#rmseVec = pandas.Series({key: None for key in lambdaParam}, name='RMSE)
N = 50
for l in lambdaParam:
measuredRMSE = []
for k in range(10):
fold = numpy.arange(k * (N), ((k + 1) * 50))
mask = trainingData.index.isin(fold)
X_t = trainingData.loc[~mask, xCols]
y_t = trainingData.loc[~mask, yCols]
X = trainingData.loc[mask, xCols]
y = trainingData.loc[mask, yCols]
#reg = linear_model.Ridge(alpha = l)
#reg.fit(X_t,y_t)
B = cr.ridge_regression(X=X_t, y=y_t, lambdaParam=l)
betas = pandas.Series(data=B.flatten(), index=xCols)
#betas = reg.coef_
yFit = pandas.Series(X.dot(betas).values.flatten(),
index=y.index,
name='yFit')
measuredRMSE.append(
numpy.sqrt(numpy.mean((y.iloc[:, 0] - yFit)**2)))
rmseVec[str(l)] = numpy.mean(measuredRMSE)
print(rmseVec)
rmseVec.to_csv(os.path.join(cd.data_dir(), cd.DataSets.EX1.value,
'__sample.csv'),
index=False)
def __init__(self, x, y):
GameRect.__init__(self, x, y, 63, 74)
#self.color = pygame.Color('orange')
self.walk_force = 0
self.image = [image.load(os.path.join(data_dir(), 'art', 'orange', 'owange.png')).convert_alpha()]
self.animation = image.load(os.path.join(data_dir(), 'art', 'orange', 'orange_splat_small.png')).convert_alpha() #spritesheet.load_strip('orange_splat.png', 1362, colorkey = None)[0]
self.in_air = True
self.allowed_glide = 2
self.role = "Owange"
self.status = None
self.deadtime = 0
# owange collides with ground and boughs
self.layers = LayerType.PLAYER | LayerType.ITEMS
self.layers = 1
def read_data():
X_test = pandas.read_csv(os.path.join(dt.data_dir(), 'task1',
'X_test.csv'),
header=0,
index_col=0)
X_train = pandas.read_csv(os.path.join(dt.data_dir(), 'task1',
'X_train.csv'),
header=0,
index_col=0)
y_train = pandas.read_csv(os.path.join(dt.data_dir(), 'task1',
'y_train.csv'),
header=0,
index_col=0)
return X_test, X_train, y_train
def main(window, handle_events):
## window = Window()
## window.init()
world = World()
world.stage = 1
p1 = Platform(600, 300, 400, 50)
world.add_item(p1)
p2 = Platform(500, 600, 800, 100)
world.add_item(p2)
""" vert order:
0 3
1 2
"""
rest_length, stiffness, damping = 200, 10, 1
spring = Spring(p1, p2, lambda vs: vs[1], lambda vs: vs[0], rest_length, stiffness, damping)
world.add_spring(spring)
spring = Spring(p1, p2, lambda vs: vs[2], lambda vs: vs[3], rest_length, stiffness, damping)
world.add_spring(spring)
spring = Spring(
p1, p2, lambda vs: (vs[1] + vs[3]) / 2, lambda vs: (vs[1] + vs[3]) / 2, rest_length, 10 * stiffness, damping
)
world.add_spring(spring)
font_path = os.path.join(data_dir(), "fonts", "vinque", "vinque.ttf")
fnt = font.Font(font_path, 36)
surface = fnt.render("The adventures of...", True, (255, 255, 255))
word = Word(p2, surface, (200, 50))
world.add_word(word)
fnt = font.Font(font_path, 48)
text = "Woger the wibbly wobbly wombat"
words = [fnt.render(word, True, (255, 255, 255)) for word in text.split()]
word_positions = ((200, 75), (500, 75), (175, 250), (350, 250), (550, 250))
for surface, position in zip(words, word_positions):
word = Word(p1, surface, position)
world.add_word(word)
fnt = font.Font(font_path, 24)
surface = fnt.render("space to continue, use awwow keys in game to weach owanges", True, (0, 0, 0))
word = Word(p2, surface, (40, 550))
world.add_word(word)
render = Render(window, world)
while True:
quit = handle_events(window, world)
if quit:
break
world.update()
render.draw_world()
display.flip()
def __init__(self, window, world):
self.window = window
self.world = world
self.camera = Camera(window)
self.facing_right = 0
#self.font = font.SysFont(None, 48)
font_path = os.path.join(data_dir(),"fonts", "vinque", "vinque.ttf")
self.font = font.Font(font_path,48)
def __init__(self, window, world):
self.window = window
self.world = world
self.camera = Camera(window)
self.facing_right = 0
#self.font = font.SysFont(None, 48)
font_path = os.path.join(data_dir(), "fonts", "vinque", "vinque.ttf")
self.font = font.Font(font_path, 48)
def read(file):
logger.info('reading [{file}]'.format(file=file))
full_file_path = os.path.join(dt.data_dir(), 'task5',
'{}'.format(file[0]))
df = pandas.read_csv(full_file_path, index_col=0, header=0)
logger.info('finished reading [{file}] of shape [{shape}]'.format(
file=file, shape=df.shape))
return df.values
def destroy(self):
self.status = "Collided"
#self.image = [image.load("data/art/orange/orange_splat_small.png").convert_alpha()] #spritesheet.load_strip('orange_splat.png', 1362, colorkey = None)[0]
self.image = [
image.load(
os.path.join(data_dir(), 'art', 'orange',
'orange_splat_small.png')).convert_alpha()
]
self.body.reset_forces()
def __init__(self, **kwargs):
clause_segmenter = kwargs.get('clause_segmenter')
if clause_segmenter is None:
self._clause_segmenter = ClauseSegmenter()
else:
self._clause_segmenter = clause_segmenter
self._clause_discarder = StartOfClauseMatcher.from_file(
data_dir('skip_rules.txt'))
self._sent = None
self._tokens = []
def __init__(self, x, y):
GameRect.__init__(self, x, y, 63, 74)
#self.color = pygame.Color('orange')
self.walk_force = 0
self.image = [
image.load(os.path.join(data_dir(), 'art', 'orange',
'owange.png')).convert_alpha()
]
self.animation = image.load(
os.path.join(data_dir(), 'art', 'orange', 'orange_splat_small.png')
).convert_alpha(
) #spritesheet.load_strip('orange_splat.png', 1362, colorkey = None)[0]
self.in_air = True
self.allowed_glide = 2
self.role = "Owange"
self.status = None
self.deadtime = 0
# owange collides with ground and boughs
self.layers = LayerType.PLAYER | LayerType.ITEMS
self.layers = 1
def __init__(self, x=63, y=74):
#x = random.randint(0,550)
#y = random.randint(0,550)
GameRect.__init__(self, x, y, 31, 75)
self.image = [image.load(os.path.join(data_dir(), 'art', 'cherry', 'cherry_small.png')).convert_alpha()]
#self.animation = image.load("data/art/orange/orange_splat_small.png").convert_alpha() #spritesheet.load_strip('orange_splat.png', 1362, colorkey = None)[0]
self.in_air = True
self.role = "Cherry"
self.status = None
self.deadtime = 0
# cherry collides with ground and boughs
self.layers = 1
def dump_to_binary(X, name):
##
logger = logging.getLogger(__name__)
full_file_path = os.path.join(dt.data_dir(), 'task5',
'{name}.p'.format(name=name))
with open(full_file_path, 'w+b') as f:
logger.info(
'writing X to binary file [{file}]'.format(file=full_file_path))
pickle.dump(X, f)
return full_file_path
def label_data():
logger = logging.getLogger(__name__)
##
logger.info('import datasets')
data_folder = os.path.join(dt.data_dir(), dt.DataSets.EX4.value)
train_labeled = pd.read_hdf(os.path.join(data_folder, "train_labeled.h5"),
"train")
train_unlabeled = pd.read_hdf(
os.path.join(data_folder, "train_unlabeled.h5"), "train")
##
logger.info('import p(y|x) model predictions')
files = os.listdir(data_folder)
wanted = [
True if file.startswith('prediction_vector_') else False
for file in files
]
fetch = list(compress(files, wanted))
container = []
result = [
container.append(
pd.read_csv(os.path.join(data_folder, file),
header=None,
index_col=0).values) for file in fetch
]
predicted_labels = np.hstack(container)
predicted_mode, count = stats.mode(predicted_labels, axis=1)
decision_vector = pd.DataFrame(np.hstack((predicted_mode, count)))
decision_vector.columns = ['class', 'count']
mask = decision_vector['count'] < 10
idx = mask[~mask].index
##
logger.info('label remaining data')
train_unlabeled_labeled = train_unlabeled.loc[idx + 9000, :]
train_unlabeled_labeled.loc[:, 'y'] = decision_vector.loc[idx,
'class'].values
# === shuffle data prior to fitting
RM = np.random.RandomState(12357)
train_labeled.index = RM.permutation(train_labeled.index)
full_train_data = pd.concat([train_labeled, train_unlabeled_labeled],
axis=0)
return full_train_data
def __init__(self, x=63, y=74):
#x = random.randint(0,550)
#y = random.randint(0,550)
GameRect.__init__(self, x, y, 31, 75)
self.image = [
image.load(
os.path.join(data_dir(), 'art', 'cherry',
'cherry_small.png')).convert_alpha()
]
#self.animation = image.load("data/art/orange/orange_splat_small.png").convert_alpha() #spritesheet.load_strip('orange_splat.png', 1362, colorkey = None)[0]
self.in_air = True
self.role = "Cherry"
self.status = None
self.deadtime = 0
# cherry collides with ground and boughs
self.layers = 1
def load_mini_h5():
h5py = pytest.importorskip("h5py")
with h5py.File(data_dir('test', 'tree.mini.h5'), 'r') as io:
y = io['y'][:]
parent = io['parent'][:]
xt = io['xt'][:]
lam = io['lam'][()]
root = int(io.attrs['root'])
y = y.reshape(-1)
n = len(y)
parent = parent.astype(np.int32)
assert len(parent) == n
if parent.min() > 0:
parent -= parent.min()
root -= 1
assert parent[root] == root
assert sys.getrefcount(parent) <= 2
assert sys.getrefcount(y) <= 2
return y, parent, xt, lam, root
def play_music(self, musicname, loop=-1):
""" plays a music track. Only one can be played at a time.
So if there is one playing, it will be stopped and the new
one started.
"""
music = mixer.music
if not music: return
if music.get_busy():
#we really should fade out nicely and
#wait for the end music event, for now, CUT
music.stop()
fullname = os.path.join(data_dir(), 'music', musicname)
fullname_ogg = fullname + ".ogg"
if os.path.exists(fullname_ogg):
fullname = fullname_ogg
music.load(fullname)
music.play(loop)
music.set_volume(1.0)
def load_image(filename, colorkey=None):
""" colorkey - if -1, then use the top left pixel as the color.
colorkey - if None, then use per pixel alpha images.
"""
if filename not in IMAGE_CACHE:
if os.path.exists(filename):
fname = filename
else:
fname = os.path.join(data_dir(), "art", filename)
img = pygame.image.load(fname)
if colorkey is not None:
# color key images
if colorkey is -1:
colorkey = img.get_at((0,0))
img = img.convert()
img.set_colorkey(colorkey, RLEACCEL)
else:
# per pixel alpha images.
img = img.convert_alpha()
IMAGE_CACHE[filename] = img
return IMAGE_CACHE[filename]
def handle_events(window, world):
quit = False
for e in event.get():
if e.type == QUIT:
quit = True
break
elif e.type == KEYDOWN:
if 1 and e.key == K_s and e.mod & KMOD_SHIFT:
pygame.image.save(pygame.display.get_surface(), "screeny.png")
if e.key == K_ESCAPE:
quit = True
break
elif e.key == K_RETURN and e.mod & KMOD_ALT:
window.toggle_fullscreen()
if world.stage == 1:
#any key quits the intro
quit = True
return quit
elif world.stage == 3:
# anykey replays game, esc quits
if quit:
return quit
else:
# non-esc keypressed
return 2
# Woger
## elif woger.allowed_glide or not woger.in_air:
woger = world.player_character
if e.key == K_LEFT:
woger.do_walk(-1)
elif e.key == K_RIGHT:
woger.do_walk(1)
elif e.key == K_SPACE or e.key == K_UP:
woger.jump()
elif e.key == K_DOWN:
woger.dive()
elif world.stage in (1, 3):
# no key down, but stop here, we're in intro/outro
return quit
## elif woger.allowed_glide or not woger.in_air:
elif e.type == KEYUP:
woger = world.player_character
if e.key == K_LEFT:
woger.end_walk()
elif e.key == K_RIGHT:
woger.end_walk()
elif e.type == CLEANUP:
world.remove_collided()
elif e.type == TICK_TOCK:
world.tick()
elif e.type == ADDCHERRY:
bounds = window.width
world.add_cherry(
random.randint(-bounds / 2, bounds / 2),
random.randint(window.height - 300, window.height))
elif e.type == ADDOWANGE:
bounds = window.width
world.add_owange(
random.randint(-bounds / 2, bounds / 2),
random.randint(window.height - 300, window.height))
elif e.type == BIRDY:
bird_files = glob.glob(
os.path.join(data_dir(), 'sounds', 'birds*.ogg'))
bsounds = [os.path.basename(b[:-4]) for b in bird_files]
the_sound = random.choice(bsounds)
Sounds.sounds.play(the_sound)
return quit
"""
import os
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import data as cd
from sklearn import datasets, metrics
from sklearn.metrics import mean_squared_error, r2_score
from sklearn.linear_model import LinearRegression
from sklearn.linear_model import Perceptron
from sklearn.metrics import classification_report, confusion_matrix
from sklearn.model_selection import GridSearchCV
# Load the datasets
trainSet = pd.read_hdf(os.path.join(cd.data_dir(), cd.DataSets.EX3.value,
'train.h5'),
key='train')
testSet = pd.read_hdf(os.path.join(cd.data_dir(), cd.DataSets.EX3.value,
'test.h5'),
key='test')
sampleSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX3.value,
'sample.csv'),
header=0,
index_col=0,
float_precision='round_trip')
# Now apply the transformations to the data:
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
# Fit only to the training data
def main():
# === Read DELAYS and WEATHER data
delays = dt.get_lineie_69_data()
weather = dt.get_weather_data()
# === Focus on BUS 69
mask = delays.linie == 69
delays = delays[mask]
delays.reset_index(drop=True, inplace=True)
# === Extract exact time delays
delays.loc[:, 'diff'] = delays.ist_an_von - delays.soll_an_von
delays.loc[:, 'time'] = pandas.to_datetime(delays.soll_an_von.copy().astype(float), errors='coerce', unit='s')
delays.time = delays.time.dt.strftime('%H:%M')
delays.loc[:, 'datetime'] = pandas.to_datetime(delays.datum_von.astype(str) + ' ' + delays.time)
delays.datetime = delays.datetime.dt.round('60min')
# === Show delay pattern as a function of time of day
temp = delays.copy()
temp.loc[:, 'hour_of_day'] = pandas.to_datetime(temp.time).dt.hour
temp = temp.groupby('hour_of_day').mean()
fig, ax = plt.subplots(1)
ax.plot(temp.index, temp['diff'])
ax.set_ylabel('Average delay [s]')
ax.set_xlabel('Time of Day [HH:MM]')
for tick in ax.get_xticklabels():
tick.set_rotation(90)
plt.savefig('delay_vs_time-of-day.png')
# === Merge with WEATHER data
weatherDelays = weather.merge(delays, right_on='datetime', left_index=True, how='inner')
weatherDelays.to_csv(os.path.join(dt.data_dir(), 'weather_delays_merged.csv'))
# ==== Remove NaN where there is no public transport data
mask = weatherDelays.datetime > datetime.datetime(2018,2,4)
weatherDelays = weatherDelays[mask]
del mask
cumulativeWeatherDelays = weatherDelays.groupby('datetime').mean()
averageWeatherDelays = weatherDelays.groupby('datetime').mean()
# === Estimate DAILY SEASONALITY using Fourier transform
'''
Description:
Fourier transform of time-series data in time domain (yt, xt) to frequency domain (yf, xf):
Arguments:
:param n: (float) number of data points / observations
:param T: (float) maximum frequency of data i.e. 1H, 1m, 1s
'''
n, m = cumulativeWeatherDelays.shape
T = 1/n
yf = scipy.fftpack.fft(weatherDelays['diff'].values)
xf = numpy.linspace(0, 1/2.0 * T, n/2)
fig, ax = plt.subplots()
ax.plot(xf, 2.0/n * numpy.abs(yf[:n//2]))
# TODO Buiild function that 1. Picks data window, 2. FFT, 3. Removes frequency in xf domain, 4. IFFT, 5. Corr
# === Try to remove DAILY SEASONALITY by subtracting previous day's value
timeDelta = datetime.timedelta(days=1)
temp = cumulativeWeatherDelays.copy() - cumulativeWeatherDelays.shift(freq=timeDelta)
dailySeasoned = temp.dropna(how='all', axis=0)
dailySeasoned = dailySeasoned.interpolate()
del timeDelta, temp
plt.figure()
dailySeasoned['diff'].plot()
# === Try to remove DAILY SEASONALITY by subtracting previous weeks's value
timeDelta = datetime.timedelta(days=7)
temp = cumulativeWeatherDelays.copy() - cumulativeWeatherDelays.shift(freq=timeDelta)
weeklySeasoned = temp.dropna(how='all', axis=0)
weeklySeasoned = weeklySeasoned.interpolate()
del timeDelta, temp
plt.figure()
weeklySeasoned['diff'].plot()
# === Plot data with and without seasoning treatment
fig, axes = plt.subplots(2, sharex=True, figsize=(15, 10))
matplotlib.rc('xtick', labelsize=24)
matplotlib.rc('ytick', labelsize=24)
axis=0
axes[axis].plot(weeklySeasoned.index, cumulativeWeatherDelays.reindex(weeklySeasoned.index)['diff'])
axes[axis].set_xlabel('Without de-seasoning')
axes[axis].set_ylabel('Delay [s]')
axis+=1
axes[axis].plot(weeklySeasoned.index, weeklySeasoned['diff'])
axes[axis].set_xlabel('With de-seasoning')
axes[axis].set_ylabel('Delay [s]')
fig.savefig('seasoned_vs_de-seasoned_delay_data.png')
# === Plot data without de-seasoning and rainfall data
fig, axes = plt.subplots(2, sharex=True, figsize=(15, 10))
axis=0
axes[axis].plot(cumulativeWeatherDelays.index, cumulativeWeatherDelays['diff'])
axes[axis].set_xlabel('Without de-seasoning')
axes[axis].set_ylabel('Delay [s]')
axis+=1
axes[axis].plot(weeklySeasoned.index, weeklySeasoned['niederschlag_mm'])
axes[axis].set_xlabel('Rain data')
axes[axis].set_ylabel('Rainfall [mm]')
fig.savefig('seasoned_vs_rainfall_data.png')
# === Plot delay-vs-weather graphs for de-seasoned data
'''
Description:
Scatter plot between CUMULATIVE MM RAIN and DELAYS
'''
mask = cumulativeWeatherDelays.reindex(index=weeklySeasoned.index)['niederschlag_mm'] > 0
xData = cumulativeWeatherDelays.reindex(index=weeklySeasoned.index)['niederschlag_mm'][mask]
yData = weeklySeasoned['diff'].loc[xData.index]
corrMat = numpy.corrcoef(xData, yData)
corrCoefPatch = mpatches.Patch(color='blue', label='Correlation coefficient := %.2f' %corrMat[0][1])
plt.figure()
plt.scatter(x=xData, y=yData, marker='x')
plt.xlabel('Precipitation (mm)')
plt.ylabel('De-seasoned delay (s)')
plt.legend(handles=[corrCoefPatch])
plt.tight_layout()
plt.savefig('corr_rain_vs_delay_-_with_de-seasoning.png')
del xData, yData
mask = cumulativeWeatherDelays['niederschlag_mm'] > 0
xData = cumulativeWeatherDelays['niederschlag_mm'][mask]
yData = cumulativeWeatherDelays['diff'].loc[xData.index]
corrMat = numpy.corrcoef(xData, yData)
corrCoefPatch = mpatches.Patch(color='blue', label='Correlation coefficient := %.2f' %corrMat[0][1])
plt.figure()
plt.scatter(x=xData, y=yData, marker='x')
plt.xlabel('Precipitation [mm]')
plt.ylabel('Delay [s]')
plt.legend(handles=[corrCoefPatch])
plt.tight_layout()
plt.savefig('corr_rain_vs_delay_-_no_de-seasoning.png')
del xData, yData
'''
Description:
Time-series plot between CUMULATIVE RAIN and DE-SEASONED DELAY
'''
xData = cumulativeWeatherDelays['niederschlag_mm']
print(xData)
yData = cumulativeWeatherDelays['diff']
fig, ax = plt.subplots(2, sharex=True, figsize=(15, 10))
axis=0
ax[axis].plot(yData.index, yData)
ax[axis].set_ylabel('Delay [s]')
axis+=1
ax[axis].bar(xData.index, height=xData, width=0.05, color='green')
ax[axis].set_xlabel('YYYY-MM-DD:HH')
def prepare_data(small_sample=None):
logger = logging.getLogger(__name__)
if bool(small_sample):
##
logger.debug('Read small sample data')
data_dir = os.path.join(dt.data_dir(), dt.Tasks.TASK3.value)
x_train = pandas.read_csv(os.path.join(data_dir, 'X_train.csv'),
header=0,
index_col=0,
nrows=small_sample)
x_test = pandas.read_csv(os.path.join(data_dir, 'X_test.csv'),
header=0,
index_col=0,
nrows=small_sample)
y_train = pandas.read_csv(os.path.join(data_dir, 'y_train.csv'),
header=0,
index_col=0,
nrows=small_sample)
else:
##
logger.debug('Read data')
data_dir = os.path.join(dt.data_dir(), dt.Tasks.TASK3.value)
x_train = pandas.read_csv(os.path.join(data_dir, 'X_train.csv'),
header=0,
index_col=0)
x_test = pandas.read_csv(os.path.join(data_dir, 'X_test.csv'),
header=0,
index_col=0)
y_train = pandas.read_csv(os.path.join(data_dir, 'y_train.csv'),
header=0,
index_col=0)
##
logger.debug('Fourier transform data')
freq_x_train_matrix, pow_x_train = fourier.prepare_frequency_data_set(
X=x_train,
sample_frequency=Frequency.HERTZ.value,
low_cut=0.67,
high_cut=25,
normalize=True,
scale=True)
freq_x_test_matrix, pow_x_test = fourier.prepare_frequency_data_set(
X=x_test,
sample_frequency=Frequency.HERTZ.value,
low_cut=0.67,
high_cut=25,
normalize=True,
scale=True)
assert pow_x_test == pow_x_train
##
freq_x_train = pandas.DataFrame(freq_x_train_matrix)
freq_x_test = pandas.DataFrame(freq_x_test_matrix)
return x_train, x_test, y_train, freq_x_test, freq_x_train
def main():
# Load the datasets
trainSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX2.value,
'train.csv'),
header=0,
index_col=0,
float_precision='round_trip')
testSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX2.value,
'test.csv'),
header=0,
index_col=0,
float_precision='round_trip')
sampleSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX2.value,
'sample.csv'),
header=0,
index_col=0,
float_precision='round_trip')
xColumns = ['x' + str(i + 1) for i in range(16)]
yColumns = ['y']
# Now apply the transformations to the data:
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
scaler.fit(trainSet.loc[:, xColumns])
scaledTrainSet = scaler.transform(trainSet.loc[:, xColumns])
scaledTestSet = scaler.transform(testSet.loc[:, xColumns])
# set up SVM classifier
svm = SVC()
paramters = {'C': np.arange(1.85, 1.89, 0.001), 'kernel': ['rbf', 'poly']}
optSVM = GridSearchCV(svm, paramters)
optSVM.fit(scaledTrainSet, trainSet.loc[:, yColumns].as_matrix().flatten())
# MLP TRAINING
optSVM.fit(scaledTrainSet, trainSet.loc[:, yColumns].as_matrix().flatten())
# prediction in-sample
yOpt_in_sample = optSVM.predict(scaledTrainSet)
# classification statistics
dim = len(yOpt_in_sample)
inSampleConfusionMatrix = confusion_matrix(trainSet.loc[:, yColumns],
yOpt_in_sample)
accuracy = np.sum(np.diag(inSampleConfusionMatrix)) / dim
print("Using optimal parameter alpha, model accuracy %.4f " % accuracy)
# prediction
y_pred = optSVM.predict(scaledTestSet)
# write to pandas Series object
yPred = pd.DataFrame(y_pred, index=testSet.index, columns=['y'])
yPred.to_csv(
os.path.join(cd.data_dir(), cd.DataSets.EX2.value,
'svm_classifier_python.csv'))
# print out results of GridSearchCV
print(pd.DataFrame.from_dict(optSVM.cv_results_))
import os import matplotlib.pyplot as plt import numpy as np import pandas as pd import data as cd from sklearn import datasets, metrics from sklearn.metrics import mean_squared_error, r2_score from sklearn.linear_model import LinearRegression from sklearn.linear_model import Perceptron from sklearn.metrics import classification_report,confusion_matrix from sklearn.model_selection import GridSearchCV # Load the datasets trainSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX2.value, 'train.csv'), header=0, index_col=0, float_precision='round_trip') testSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX2.value, 'test.csv'), header=0, index_col=0, float_precision='round_trip') sampleSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX2.value, 'sample.csv'), header=0, index_col=0, float_precision='round_trip') # Now apply the transformations to the data: from sklearn.preprocessing import StandardScaler scaler = StandardScaler() # Fit only to the training data prcptrn = Perceptron(fit_intercept=False) xColumns = ['x' + str(i+1) for i in range(16)] yColumns = ['y'] scaler.fit(trainSet.loc[:, xColumns]) scaledTrainSet = scaler.transform(trainSet.loc[:, xColumns])
ch = logging.StreamHandler(sys.stdout)
ch.setLevel(logging.DEBUG)
formatter = logging.Formatter(
'%(asctime)s - %(name)s - %(levelname)s - %(message)s')
ch.setFormatter(formatter)
logger.addHandler(ch)
import fourier
import arima
import bios
#nrows = 10
nrows = 3030 + 443 + 1474 + 170
regenerate = False
y = pd.read_csv(os.path.join(dt.data_dir(), 'task3', 'y_train.csv'),
header=0,
index_col=0,
nrows=nrows)
_X = pd.read_csv(os.path.join(dt.data_dir(), 'task3', 'X_train.csv'),
header=0,
index_col=0,
nrows=nrows)
_X_test = pd.read_csv(os.path.join(dt.data_dir(), 'task3', 'X_test.csv'),
header=0,
index_col=0,
nrows=nrows)
if regenerate:
X_templates = bios.templates(_X)
X_test_templates = bios.templates(_X_test)
def main():
# === Data
train = pandas.read_hdf(os.path.join(cd.data_dir(), cd.DataSets.EX3.value,
'train.h5'),
key='train')
test = pandas.read_hdf(os.path.join(cd.data_dir(), cd.DataSets.EX3.value,
'test.h5'),
key='test')
train_tensor = tf.convert_to_tensor(train.as_matrix())
batches = tf.split(train, num_or_size_splits=36, axis=0)
print(batches)
features = ["x%i" % i for i in numpy.arange(1, 101, 1)]
target = ['y']
n, F = train[features].shape
C = train[target].unique
# ==== Parameters
learning_rate = 0.1
num_steps = 500
batch_size = 128
display_step = 100
# === Network Parameters
n_hidden_1 = 256 # 1st layer number of neurons
n_hidden_2 = 256 # 2nd layer number of neurons
num_input = F
num_classes = len(C) # MNIST total classes (0-9 digits)
# === tf Graph input - reserve space / decalre
X = tf.placeholder("float", [None, num_input])
Y = tf.placeholder("float", [None, num_classes])
# === Store layers weight & bias
weights = {
'h1': tf.Variable(tf.random_normal([num_input, n_hidden_1])),
'h2': tf.Variable(tf.random_normal([n_hidden_1, n_hidden_2])),
'out': tf.Variable(tf.random_normal([n_hidden_2, num_classes]))
}
biases = {
'b1': tf.Variable(tf.random_normal([n_hidden_1])),
'b2': tf.Variable(tf.random_normal([n_hidden_2])),
'out': tf.Variable(tf.random_normal([num_classes]))
}
# === Construct model
logits = neural_net(X)
prediction = tf.nn.softmax(logits)
# Define loss and optimizer
loss_op = tf.reduce_mean(
tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=Y))
optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)
train_op = optimizer.minimize(loss_op)
# Evaluate model
correct_pred = tf.equal(tf.argmax(prediction, 1), tf.argmax(Y, 1))
accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
# Initialize the variables (i.e. assign their default value)
init = tf.global_variables_initializer()
# Start training
with tf.Session() as sess:
# Run the initializer
sess.run(init)
for step in range(1, num_steps + 1):
for batch in batches:
batch_x, batch_y = mnist.train.next_batch(batch_size)
# Run optimization op (backprop)
sess.run(train_op, feed_dict={X: batch_x, Y: batch_y})
if step % display_step == 0 or step == 1:
# Calculate batch loss and accuracy
loss, acc = sess.run([loss_op, accuracy],
feed_dict={
X: batch_x,
Y: batch_y
})
print("Step " + str(step) + ", Minibatch Loss= " + \
"{:.4f}".format(loss) + ", Training Accuracy= " + \
"{:.3f}".format(acc))
print("Optimization Finished!")
# Calculate accuracy for MNIST test images
print("Testing Accuracy:", \
sess.run(accuracy, feed_dict={X: mnist.test.images,
Y: mnist.test.labels}))
return {}
import sys
import logging
logger = logging.getLogger(__name__)
logger.setLevel(logging.INFO)
ch = logging.StreamHandler(sys.stdout)
ch.setLevel(logging.DEBUG)
formatter = logging.Formatter(
'%(asctime)s - %(name)s - %(levelname)s - %(message)s')
ch.setFormatter(formatter)
logger.addHandler(ch)
#nrows = 10
nrows = 3030 + 443 + 1474 + 170
y = pd.read_csv(os.path.join(dt.data_dir(), 'task3', 'y_train.csv'),
header=0,
index_col=0,
nrows=nrows)
_X = pd.read_csv(os.path.join(dt.data_dir(), 'task3', 'X_train.csv'),
header=0,
index_col=0,
nrows=nrows)
_X_test = pd.read_csv(os.path.join(dt.data_dir(), 'task3', 'X_test.csv'),
header=0,
index_col=0,
nrows=nrows)
if True:
X_templates = templates(_X)
X_test_templates = templates(_X_test)
## print ("total_width is:%s:" % total_width)
big_surf = pygame.Surface((total_width, surfs[0].get_height()), SRCALPHA, 32)
x = 0
for s in surfs:
big_surf.blit(s, (x, 0))
x += surfs[0].get_width()
pygame.image.save(big_surf, out_fname)
if __name__ == "__main__":
if 1:
combine_images_into_sprite_sheet(os.path.join(data_dir(), 'art', 'leaves'),
os.path.join(data_dir(), 'art', 'leaves-rotating-88.png'))
else:
pygame.init()
screen = pygame.display.set_mode((640,480))
# here we just load the strip into a whole bunch of sub surfaces.
# sub surfaces reference the big image, but act just like normal surfaces.
sub_surfaces = load_strip('leaf-movement-88.png', 88, colorkey = None)
## print len(sub_surfaces)
assert(len(sub_surfaces) == 2)
@author: Anastasiya
"""
import os
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import data as cd
from sklearn import datasets, metrics
from sklearn.metrics import mean_squared_error, r2_score
from sklearn.linear_model import LinearRegression
# Load the datasets
trainSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX0.value,
'train.csv'),
header=0,
index_col=0,
float_precision='round_trip')
testSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX0.value,
'test.csv'),
header=0,
index_col=0,
float_precision='round_trip')
sampleSet = pd.read_csv(os.path.join(cd.data_dir(), cd.DataSets.EX0.value,
'sample.csv'),
header=0,
index_col=0,
float_precision='round_trip')
def main(window, handle_events):
## window = Window()
## window.init()
world = World()
world.stage = 1
p1 = Platform(600, 300, 400, 50)
world.add_item(p1)
p2 = Platform(500, 600, 800, 100)
world.add_item(p2)
""" vert order:
0 3
1 2
"""
rest_length, stiffness, damping = 200, 10, 1
spring = Spring(p1, p2, lambda vs: vs[1], lambda vs: vs[0], rest_length,
stiffness, damping)
world.add_spring(spring)
spring = Spring(p1, p2, lambda vs: vs[2], lambda vs: vs[3], rest_length,
stiffness, damping)
world.add_spring(spring)
spring = Spring(p1, p2, lambda vs: (vs[1] + vs[3]) / 2, lambda vs:
(vs[1] + vs[3]) / 2, rest_length, 10 * stiffness, damping)
world.add_spring(spring)
font_path = os.path.join(data_dir(), "fonts", "vinque", "vinque.ttf")
fnt = font.Font(font_path, 36)
surface = fnt.render('The adventures of...', True, (255, 255, 255))
word = Word(p2, surface, (200, 50))
world.add_word(word)
fnt = font.Font(font_path, 48)
text = 'Woger the wibbly wobbly wombat'
words = [fnt.render(word, True, (255, 255, 255)) for word in text.split()]
word_positions = (
(200, 75),
(500, 75),
(175, 250),
(350, 250),
(550, 250),
)
for surface, position in zip(words, word_positions):
word = Word(p1, surface, position)
world.add_word(word)
fnt = font.Font(font_path, 24)
surface = fnt.render(
'space to continue, use awwow keys in game to weach owanges', True,
(0, 0, 0))
word = Word(p2, surface, (40, 550))
world.add_word(word)
render = Render(window, world)
while True:
quit = handle_events(window, world)
if quit:
break
world.update()
render.draw_world()
display.flip()
def handle_events(window, world):
quit = False
for e in event.get():
if e.type == QUIT:
quit = True
break
elif e.type == KEYDOWN:
if 1 and e.key == K_s and e.mod & KMOD_SHIFT:
pygame.image.save( pygame.display.get_surface() , "screeny.png")
if e.key == K_ESCAPE:
quit = True
break
elif e.key == K_RETURN and e.mod & KMOD_ALT:
window.toggle_fullscreen()
if world.stage == 1:
#any key quits the intro
quit = True
return quit
elif world.stage == 3:
# anykey replays game, esc quits
if quit:
return quit
else:
# non-esc keypressed
return 2
# Woger
## elif woger.allowed_glide or not woger.in_air:
woger = world.player_character
if e.key == K_LEFT:
woger.do_walk(-1)
elif e.key == K_RIGHT:
woger.do_walk(1)
elif e.key == K_SPACE or e.key == K_UP:
woger.jump()
elif e.key == K_DOWN:
woger.dive()
elif world.stage in (1, 3):
# no key down, but stop here, we're in intro/outro
return quit
## elif woger.allowed_glide or not woger.in_air:
elif e.type == KEYUP:
woger = world.player_character
if e.key == K_LEFT:
woger.end_walk()
elif e.key == K_RIGHT:
woger.end_walk()
elif e.type == CLEANUP:
world.remove_collided()
elif e.type == TICK_TOCK:
world.tick()
elif e.type == ADDCHERRY:
bounds = window.width
world.add_cherry(random.randint(-bounds/2, bounds/2),
random.randint(window.height-300,window.height ))
elif e.type == ADDOWANGE:
bounds = window.width
world.add_owange(random.randint(-bounds/2, bounds/2),
random.randint(window.height-300,window.height ))
elif e.type == BIRDY:
bird_files = glob.glob(os.path.join(data_dir(),'sounds','birds*.ogg'))
bsounds = [os.path.basename(b[:-4]) for b in bird_files]
the_sound = random.choice(bsounds)
Sounds.sounds.play(the_sound)
return quit
def main():
N = 21600
validate_size = 2000
epochs = 50
type = ModelType.CNN_LSTM
###################################
### Read train data and fit models
###################################
eeg1 = pd.read_csv(os.path.join(dt.data_dir(), 'task5', 'train_eeg1.csv'),
header=0,
index_col=0)
eeg2 = pd.read_csv(os.path.join(dt.data_dir(), 'task5', 'train_eeg2.csv'),
header=0,
index_col=0)
emg = pd.read_csv(os.path.join(dt.data_dir(), 'task5', 'train_emg.csv'),
header=0,
index_col=0)
labels = pd.read_csv(os.path.join(dt.data_dir(), 'task5',
'train_labels.csv'),
header=0,
index_col=0)
##########################
### subject one model
start = 0
end = N
label = 'subject1_%s_%s_epochs' % (type, epochs)
subject1_model = train(eeg1=eeg1.iloc[start:end, :],
eeg2=eeg2.iloc[start:end, :],
emg=emg.iloc[start:end, :],
labels=labels.iloc[start:end, :],
type=type,
validate_size=validate_size,
epochs=epochs,
label=label)
##########################
### subject two model
start = N
end = N * 2
label = 'subject2_%s_%s_epochs' % (type, epochs)
subject2_model = train(eeg1=eeg1.iloc[start:end, :],
eeg2=eeg2.iloc[start:end, :],
emg=emg.iloc[start:end, :],
labels=labels.iloc[start:end, :],
type=type,
validate_size=validate_size,
epochs=epochs,
label=label)
##########################
### subject three model
start = N * 2
end = N * 3 - 500
label = 'subject3_%s_%s_epochs' % (type, epochs)
subject3_model = train(eeg1=eeg1.iloc[start:end, :],
eeg2=eeg2.iloc[start:end, :],
emg=emg.iloc[start:end, :],
labels=labels.iloc[start:end, :],
type=type,
validate_size=validate_size,
epochs=epochs,
label=label)
##############################################
### Models fitted, read test data and predict
##############################################
eeg1_test = pd.read_csv(os.path.join(dt.data_dir(), 'task5',
'test_eeg1.csv'),
header=0,
index_col=0)
eeg2_test = pd.read_csv(os.path.join(dt.data_dir(), 'task5',
'test_eeg2.csv'),
header=0,
index_col=0)
emg_test = pd.read_csv(os.path.join(dt.data_dir(), 'task5',
'test_emg.csv'),
header=0,
index_col=0)
eeg1_test_A = df_row_norm(eeg1_test.iloc[:N, :].fillna(0)).values
eeg2_test_A = df_row_norm(eeg2_test.iloc[:N, :].fillna(0)).values
emg_test_A = df_row_norm(emg_test.iloc[:N, :].fillna(0)).values
eeg1_test_B = df_row_norm(eeg1_test.iloc[N:, :].fillna(0)).values
eeg2_test_B = df_row_norm(eeg2_test.iloc[N:, :].fillna(0)).values
emg_test_B = df_row_norm(emg_test.iloc[N:, :].fillna(0)).values
X_test_A = np.dstack((eeg1_test_A, eeg2_test_A, emg_test_A))
X_test_B = np.dstack((eeg1_test_B, eeg2_test_B, emg_test_B))
#################################
### subject one model prediction
label = 'subject_1_%s_weighted_%s_epochs' % (type, epochs)
y_subject1_score = predict(X_test_A,
X_test_B,
model=subject1_model,
type=type,
weights=[1, 0.5, 2.5],
label=label)
#################################
### subject two model prediction
label = 'subject_2_%s_weighted_%s_epochs' % (type, epochs)
y_subject2_score = predict(X_test_A,
X_test_B,
model=subject2_model,
type=type,
weights=[1, 0.5, 2.0],
label=label)
###################################
### subject three model prediction
label = 'subject_3_%s_weighted_%s_epochs' % (type, epochs)
y_subject3_score = predict(X_test_A,
X_test_B,
model=subject3_model,
type=type,
weights=[1, 0.5, 4.5],
label=label)
##################################
### all subjects model prediction
label = 'all_subjects_%s_%s_epochs' % (type, epochs)
y_score = y_subject1_score * 0.33 + y_subject2_score * 0.33 + y_subject3_score * 0.33
y_test = np.argmax(y_score, axis=1)
result = pd.Series(y_test)
expected = [0.526, 0.418, 0.0548]
for i in range(3):
print("class expected/realized class ratio [%s]: [%s/%s]" %
(i, expected[i], sum(result == i) / len(result)))
print("")
result += 1
result.index.name = 'Id'
result.name = 'y'
pd.DataFrame(result).to_csv(os.path.join(dt.output_dir(),
"%s.csv" % label))
##################################
### all subjects model prediction
label = 'all_subjects_%s_weighted_%s_epochs' % (type, epochs)
y_score = (y_subject1_score * 0.5 + y_subject2_score * 0 +
y_subject3_score * 0.5) * [1.5, 0.8, 1.6]
y_test = np.argmax(y_score, axis=1)
result = pd.Series(y_test)
expected = [0.526, 0.418, 0.0548]
for i in range(3):
print("class expected/realized class ratio [%s]: [%s/%s]" %
(i, expected[i], sum(result == i) / len(result)))
print("")
result += 1
result.index.name = 'Id'
result.name = 'y'
pd.DataFrame(result).to_csv(os.path.join(dt.output_dir(),
"%s.csv" % label))
print("DONE")
def destroy(self):
self.status = "Collided"
#self.image = [image.load("data/art/orange/orange_splat_small.png").convert_alpha()] #spritesheet.load_strip('orange_splat.png', 1362, colorkey = None)[0]
self.image = [image.load(os.path.join(data_dir(), 'art', 'orange', 'orange_splat_small.png')).convert_alpha()]
self.body.reset_forces()
def main(window, handle_events, score):
## window = Window()
## window.init()
world = World()
world.stage = 3
p1 = Platform(600, 300, 400, 50)
world.add_item(p1)
p2 = Platform(500, 600, 800, 100)
world.add_item(p2)
""" vert order:
0 3
1 2
"""
rest_length, stiffness, damping = 200, 10, 1
spring = Spring(p1, p2, lambda vs: vs[1], lambda vs: vs[0], rest_length,
stiffness, damping)
world.add_spring(spring)
spring = Spring(p1, p2, lambda vs: vs[2], lambda vs: vs[3], rest_length,
stiffness, damping)
world.add_spring(spring)
spring = Spring(p1, p2, lambda vs: (vs[1] + vs[3]) / 2, lambda vs:
(vs[1] + vs[3]) / 2, rest_length, 10 * stiffness, damping)
world.add_spring(spring)
font_path = os.path.join(data_dir(), "fonts", "vinque", "vinque.ttf")
## fnt = font.Font(font_path, 36)
## surface = fnt.render('The adventures of...', True, (255,255,255))
## word = Word(p2, surface, (200, 50))
## world.add_word(word)
fnt = font.Font(font_path, 48)
text = 'You scored'
words = [fnt.render(word, True, (0, 0, 0)) for word in text.split()]
word_positions = (
(150, 60),
(550, 60),
)
for surface, position in zip(words, word_positions):
word = Word(p1, surface, position)
world.add_word(word)
fnt = font.Font(font_path, 96)
surface = fnt.render('% 2d' % score, True, (255, 255, 255))
word = Word(p2, surface, (230, 350))
world.add_word(word)
fnt = font.Font(font_path, 24)
surface = fnt.render('space to play again, escape to quit', True,
(0, 0, 0))
word = Word(p2, surface, (200, 550))
world.add_word(word)
render = Render(window, world)
while True:
quit = handle_events(window, world)
if quit == 2:
# non-esc keypressed
return False
if quit:
return quit
world.update()
render.draw_world()
display.flip()
def main(window, handle_events, score):
## window = Window()
## window.init()
world = World()
world.stage = 3
p1 = Platform(600, 300, 400, 50)
world.add_item(p1)
p2 = Platform(500, 600, 800, 100)
world.add_item(p2)
""" vert order:
0 3
1 2
"""
rest_length, stiffness, damping = 200, 10, 1
spring = Spring(p1, p2,
lambda vs:vs[1],
lambda vs:vs[0],
rest_length, stiffness, damping)
world.add_spring(spring)
spring = Spring(p1, p2,
lambda vs:vs[2],
lambda vs:vs[3],
rest_length, stiffness, damping)
world.add_spring(spring)
spring = Spring(p1, p2,
lambda vs: (vs[1] + vs[3])/2,
lambda vs: (vs[1] + vs[3])/2,
rest_length, 10*stiffness, damping)
world.add_spring(spring)
font_path = os.path.join(data_dir(), "fonts", "vinque", "vinque.ttf")
## fnt = font.Font(font_path, 36)
## surface = fnt.render('The adventures of...', True, (255,255,255))
## word = Word(p2, surface, (200, 50))
## world.add_word(word)
fnt = font.Font(font_path, 48)
text = 'You scored'
words = [fnt.render(word, True, (0,0,0))
for word in text.split()]
word_positions = (
(150, 60),
(550, 60),
)
for surface, position in zip(words, word_positions):
word = Word(p1, surface, position)
world.add_word(word)
fnt = font.Font(font_path, 96)
surface = fnt.render('% 2d' %score, True, (255,255,255))
word = Word(p2, surface, (230, 350))
world.add_word(word)
fnt = font.Font(font_path, 24)
surface = fnt.render('space to play again, escape to quit', True, (0,0,0))
word = Word(p2, surface, (200, 550))
world.add_word(word)
render = Render(window, world)
while True:
quit = handle_events(window, world)
if quit == 2:
# non-esc keypressed
return False
if quit:
return quit
world.update()
render.draw_world()
display.flip()
from sklearn.metrics import mean_squared_error
from sklearn.model_selection import train_test_split
from sklearn.naive_bayes import GaussianNB
import logging
import tensorflow as tf
# hello = tf.constant('Hello tf!')
# sess = tf.Session()
# print(sess.run(hello))
log = logging.getLogger(__name__)
#data_folder = '/Users/dmitrykazakov/Desktop/Studium/MSc/2. Semester/ML/projects/task4_s8n2k3nd/data/ex4/'
data_folder = os.path.join(dt.data_dir(), dt.DataSets.EX4.value)
def input_eval_set(feature_names, feature_values, class_labels=None):
features = dict(
zip(feature_names,
np.matrix(feature_values).transpose().tolist()))
if class_labels is None:
return features
else:
labels = np.int32(np.array(class_labels))
return features, labels
def train_input_fn(features, labels, batch_size):
NOTE: not using pygames channel queueing as it only allows one sound to be
queued. Also the sound can only be queued on a certain channel.
"""
from pygame import mixer
import os
import glob
#the as alias allows re-imports
from cyclic_list import cyclic_list as cyclic_list_func
from data import data_dir
SOUND_PATH = os.path.join(data_dir(), "sounds")
def get_sound_list(path = SOUND_PATH):
""" gets a list of sound names without thier path, or extension.
"""
# load a list of sounds without path at the beginning and .ogg at the end.
sound_list = []
for ext in ['.wav', '.ogg']:
sound_list += map(lambda x:x[len(path)+1:-4],
#glob.glob(os.path.join(path,"*.ogg"))
glob.glob(os.path.join(path,"*" + ext))
)
return sound_list