def __startRecordingAndStreaming():
# Connect a client socket to my_server:8000 (change my_server to the
# hostname of your server)
if isMonitoringWorking():
return
global __monitoringWorking
__monitoringWorking = True
LedChanger.lightPhotoLedOn()
print("Monitoring has been started")
try:
with picamera.PiCamera() as camera:
global __camera
__camera = camera
camera.resolution = (1640, 1232)
# camera.framerate = 23
import DataManager
from DataManager import deviceName, videoDir
DataManager.createVideoDirIfNotExists()
videoPath = str(videoDir) + str(deviceName) + "_" + str(
datetime.datetime.now()) + '.h264'
print("a")
camera.start_recording(videoPath, resize=(1024, 768))
print("b")
__currentMonitoringPeriodicTask.launchMonitoringPeriodicTask(
camera, videoPath)
except Exception as e:
utils.printException(e)
onMonitoringStopped()
print("__startRecordingAndStreaming finished")
def requestGetFromAliAPI(url, successBlock=None, failureBlock=None):
global m_ctx
bodys = {}
method = 'GET'
# 拼接url
Logger.log("now request.url = " + url)
urlReq = request.Request(url)
urlReq.add_header('Authorization', 'APPCODE ' + constant.AppCode)
response = request.urlopen(urlReq, context=m_ctx)
respData = response.read()
# respData = "{\"showapi_res_code\":0,\"showapi_res_error\":\"\",\"showapi_res_body\":{\"ret_code\":0,\"list\":[{\"trade_money\":\"368947118.000\",\"diff_money\":\"-0.03\",\"open_price\":\"8.880\",\"code\":\"601006\",\"date\":\"2017-11-30\",\"min_price\":\"8.810\",\"market\":\"sh\",\"trade_num\":\"416807\",\"turnover\":\"0.280\",\"close_price\":\"8.850\",\"max_price\":\"8.910\",\"swing\":\"1.13\",\"diff_rate\":\"-0.34\"},{\"trade_money\":\"388937580.000\",\"diff_money\":\"-0.02\",\"open_price\":\"8.900\",\"code\":\"601006\",\"date\":\"2017-11-29\",\"min_price\":\"8.820\",\"market\":\"sh\",\"trade_num\":\"438484\",\"turnover\":\"0.295\",\"close_price\":\"8.880\",\"max_price\":\"8.920\",\"swing\":\"1.12\",\"diff_rate\":\"-0.22\"},{\"trade_money\":\"338012785.000\",\"diff_money\":\"-0.12\",\"open_price\":\"8.940\",\"code\":\"601006\",\"date\":\"2017-11-28\",\"min_price\":\"8.870\",\"market\":\"sh\",\"trade_num\":\"379584\",\"turnover\":\"0.255\",\"close_price\":\"8.900\",\"max_price\":\"8.970\",\"swing\":\"1.11\",\"diff_rate\":\"-1.33\"},{\"trade_money\":\"679468677.000\",\"diff_money\":\"0.0\",\"open_price\":\"8.940\",\"code\":\"601006\",\"date\":\"2017-11-27\",\"min_price\":\"8.790\",\"market\":\"sh\",\"trade_num\":\"762013\",\"turnover\":\"0.513\",\"close_price\":\"9.020\",\"max_price\":\"9.040\",\"swing\":\"2.77\",\"diff_rate\":\"0.0\"},{\"trade_money\":\"716821338.000\",\"diff_money\":\"-0.09\",\"open_price\":\"9.060\",\"code\":\"601006\",\"date\":\"2017-11-24\",\"min_price\":\"8.810\",\"market\":\"sh\",\"trade_num\":\"801930\",\"turnover\":\"0.539\",\"close_price\":\"9.020\",\"max_price\":\"9.100\",\"swing\":\"3.18\",\"diff_rate\":\"-0.99\"},{\"trade_money\":\"1061585849.000\",\"diff_money\":\"-0.27\",\"open_price\":\"9.400\",\"code\":\"601006\",\"date\":\"2017-11-23\",\"min_price\":\"9.030\",\"market\":\"sh\",\"trade_num\":\"1141401\",\"turnover\":\"0.768\",\"close_price\":\"9.110\",\"max_price\":\"9.490\",\"swing\":\"4.9\",\"diff_rate\":\"-2.88\"},{\"trade_money\":\"930883985.000\",\"diff_money\":\"0.09\",\"open_price\":\"9.310\",\"code\":\"601006\",\"date\":\"2017-11-22\",\"min_price\":\"9.210\",\"market\":\"sh\",\"trade_num\":\"1001580\",\"turnover\":\"0.674\",\"close_price\":\"9.380\",\"max_price\":\"9.390\",\"swing\":\"1.94\",\"diff_rate\":\"0.97\"},{\"trade_money\":\"1164999730.000\",\"diff_money\":\"0.26\",\"open_price\":\"9.000\",\"code\":\"601006\",\"date\":\"2017-11-21\",\"min_price\":\"8.960\",\"market\":\"sh\",\"trade_num\":\"1268321\",\"turnover\":\"0.853\",\"close_price\":\"9.290\",\"max_price\":\"9.350\",\"swing\":\"4.32\",\"diff_rate\":\"2.88\"},{\"trade_money\":\"471383497.000\",\"diff_money\":\"0.0\",\"open_price\":\"9.000\",\"code\":\"601006\",\"date\":\"2017-11-20\",\"min_price\":\"8.890\",\"market\":\"sh\",\"trade_num\":\"525579\",\"turnover\":\"0.354\",\"close_price\":\"9.030\",\"max_price\":\"9.060\",\"swing\":\"1.88\",\"diff_rate\":\"0.0\"},{\"trade_money\":\"971205411.000\",\"diff_money\":\"0.24\",\"open_price\":\"8.800\",\"code\":\"601006\",\"date\":\"2017-11-17\",\"min_price\":\"8.700\",\"market\":\"sh\",\"trade_num\":\"1093673\",\"turnover\":\"0.736\",\"close_price\":\"9.030\",\"max_price\":\"9.060\",\"swing\":\"4.1\",\"diff_rate\":\"2.73\"},{\"trade_money\":\"285205565.000\",\"diff_money\":\"-0.12\",\"open_price\":\"8.890\",\"code\":\"601006\",\"date\":\"2017-11-16\",\"min_price\":\"8.780\",\"market\":\"sh\",\"trade_num\":\"323445\",\"turnover\":\"0.218\",\"close_price\":\"8.790\",\"max_price\":\"8.890\",\"swing\":\"1.23\",\"diff_rate\":\"-1.35\"},{\"trade_money\":\"352421722.000\",\"diff_money\":\"-0.04\",\"open_price\":\"8.930\",\"code\":\"601006\",\"date\":\"2017-11-15\",\"min_price\":\"8.830\",\"market\":\"sh\",\"trade_num\":\"396352\",\"turnover\":\"0.267\",\"close_price\":\"8.910\",\"max_price\":\"8.960\",\"swing\":\"1.45\",\"diff_rate\":\"-0.45\"},{\"trade_money\":\"272978108.000\",\"diff_money\":\"-0.04\",\"open_price\":\"9.000\",\"code\":\"601006\",\"date\":\"2017-11-14\",\"min_price\":\"8.910\",\"market\":\"sh\",\"trade_num\":\"304990\",\"turnover\":\"0.205\",\"close_price\":\"8.950\",\"max_price\":\"9.010\",\"swing\":\"1.11\",\"diff_rate\":\"-0.44\"},{\"trade_money\":\"371140472.000\",\"diff_money\":\"0.15\",\"open_price\":\"8.840\",\"code\":\"601006\",\"date\":\"2017-11-13\",\"min_price\":\"8.820\",\"market\":\"sh\",\"trade_num\":\"415312\",\"turnover\":\"0.279\",\"close_price\":\"8.990\",\"max_price\":\"9.020\",\"swing\":\"2.26\",\"diff_rate\":\"1.7\"},{\"trade_money\":\"332588918.000\",\"diff_money\":\"-0.08\",\"open_price\":\"8.920\",\"code\":\"601006\",\"date\":\"2017-11-10\",\"min_price\":\"8.790\",\"market\":\"sh\",\"trade_num\":\"376272\",\"turnover\":\"0.253\",\"close_price\":\"8.840\",\"max_price\":\"8.930\",\"swing\":\"1.57\",\"diff_rate\":\"-0.9\"},{\"trade_money\":\"241490411.000\",\"diff_money\":\"0.04\",\"open_price\":\"8.870\",\"code\":\"601006\",\"date\":\"2017-11-09\",\"min_price\":\"8.850\",\"market\":\"sh\",\"trade_num\":\"271364\",\"turnover\":\"0.183\",\"close_price\":\"8.920\",\"max_price\":\"8.930\",\"swing\":\"0.9\",\"diff_rate\":\"0.45\"},{\"trade_money\":\"377205964.000\",\"diff_money\":\"0.02\",\"open_price\":\"8.860\",\"code\":\"601006\",\"date\":\"2017-11-08\",\"min_price\":\"8.830\",\"market\":\"sh\",\"trade_num\":\"424379\",\"turnover\":\"0.285\",\"close_price\":\"8.880\",\"max_price\":\"8.960\",\"swing\":\"1.47\",\"diff_rate\":\"0.23\"},{\"trade_money\":\"404532369.000\",\"diff_money\":\"0.03\",\"open_price\":\"8.820\",\"code\":\"601006\",\"date\":\"2017-11-07\",\"min_price\":\"8.780\",\"market\":\"sh\",\"trade_num\":\"457014\",\"turnover\":\"0.307\",\"close_price\":\"8.860\",\"max_price\":\"8.900\",\"swing\":\"1.36\",\"diff_rate\":\"0.34\"},{\"trade_money\":\"326183604.000\",\"diff_money\":\"-0.01\",\"open_price\":\"8.840\",\"code\":\"601006\",\"date\":\"2017-11-06\",\"min_price\":\"8.760\",\"market\":\"sh\",\"trade_num\":\"370228\",\"turnover\":\"0.249\",\"close_price\":\"8.830\",\"max_price\":\"8.860\",\"swing\":\"1.13\",\"diff_rate\":\"-0.11\"},{\"trade_money\":\"606428072.000\",\"diff_money\":\"-0.1\",\"open_price\":\"8.940\",\"code\":\"601006\",\"date\":\"2017-11-03\",\"min_price\":\"8.720\",\"market\":\"sh\",\"trade_num\":\"685665\",\"turnover\":\"0.461\",\"close_price\":\"8.840\",\"max_price\":\"8.950\",\"swing\":\"2.57\",\"diff_rate\":\"-1.12\"},{\"trade_money\":\"322562461.000\",\"diff_money\":\"0.03\",\"open_price\":\"8.920\",\"code\":\"601006\",\"date\":\"2017-11-02\",\"min_price\":\"8.840\",\"market\":\"sh\",\"trade_num\":\"361503\",\"turnover\":\"0.243\",\"close_price\":\"8.940\",\"max_price\":\"8.980\",\"swing\":\"1.57\",\"diff_rate\":\"0.34\"},{\"trade_money\":\"682116475.000\",\"diff_money\":\"-0.19\",\"open_price\":\"9.100\",\"code\":\"601006\",\"date\":\"2017-11-01\",\"min_price\":\"8.890\",\"market\":\"sh\",\"trade_num\":\"760051\",\"turnover\":\"0.511\",\"close_price\":\"8.910\",\"max_price\":\"9.100\",\"swing\":\"2.31\",\"diff_rate\":\"-2.09\"}]}}"
if (respData):
DataManager.writeOriginalData(respData)
Logger.logTip("connect success! 返回的数据如下：\n %s" % respData)
jsonAttrs = json.loads(respData)
if jsonAttrs.get("showapi_res_code") != None and jsonAttrs.get(
"showapi_res_code") == 0:
Logger.logTip("getRequest success! url = [%s]" % url)
bodyData = jsonAttrs.get("showapi_res_body")
if (successBlock):
successBlock(bodyData)
else:
errorMsg = jsonAttrs.get("showapi_res_error")
Logger.logError("url respone errorcode! ErrorMsg = %s" % errorMsg)
if (failureBlock):
failureBlock(errorMsg)
else:
Logger.logError("网络链接错误！url = %s" % url)
def CPConvergenceTest():
value = []
error = []
eventFiles = dm.GetFileNames('\Samples') # get filenames
eventFiles_CP = dm.GetFileNames('\Samples_CP')
for i in range(len(eventFiles)):
p = dm.AmpGendf(eventFiles[i], False) # generate particle data
pbar = dm.AmpGendf(eventFiles_CP[i], True) # generate CP particle data
C_T = kin.Scalar_TP(kin.Vector_3(p['p_3']), kin.Vector_3(p['p_4']),
kin.Vector_3(
p['p_1'])) # calcualtes scalar triple product
C_Tbar = -kin.Scalar_TP(
kin.Vector_3(pbar['p_3']), kin.Vector_3(pbar['p_4']),
kin.Vector_3(pbar['p_1'])) # -sign for parity flip
A_T = kin.TP_Amplitude(C_T) # calculate parity asymmetries
A_Tbar = kin.TP_Amplitude(C_Tbar)
A_CP = kin.A_CP(A_T, A_Tbar) # calculate A_CP
value.append(A_CP[0])
error.append(A_CP[1])
pt.ErrorPlot([np.linspace(1, 10, len(eventFiles)), value],
axis=True,
y_error=error,
x_axis="Number of Events ($10^{5}$)",
y_axis="$\mathcal{A}_{CP}$") # plots data
def Seed_test():
fileNames = dm.GetFileNames('\seed_test') # get filenames
events = fileNames[0:5] # split the dataset in half
events_CP = fileNames[5:10] # make this half CP data
value = []
error = []
for i in range(5):
p = dm.AmpGendf(events[i], False) # generate particle data
pbar = dm.AmpGendf(events_CP[i], True)
C_T = kin.Scalar_TP(kin.Vector_3(p['p_3']), kin.Vector_3(p['p_4']),
kin.Vector_3(
p['p_1'])) # calcualtes scalar triple product
C_Tbar = -kin.Scalar_TP(kin.Vector_3(
pbar['p_3']), kin.Vector_3(pbar['p_4']), kin.Vector_3(pbar['p_1']))
A_T = kin.TP_Amplitude(C_T) # calculate parity asymmetries
A_Tbar = kin.TP_Amplitude(C_Tbar)
A_CP = kin.A_CP(A_T, A_Tbar) # calculate A_CP
value.append(A_CP[0])
error.append(A_CP[1])
pt.ErrorPlot([np.linspace(1, 5, 5), value],
axis=True,
y_error=error,
x_axis="Iteration",
y_axis="$\mathcal{A}_{CP}$") # plots data
def __init__(self, trackFile, truthFile):
self.trackManager = dm.DataManager(trackFile, hasId=True)
self.truthManager = dm.DataManager(truthFile, hasId=True)
self.IMPOSSIBLE_SCORE = 1.0E10
self.NEW_TRACK_SCORE = 10.0
self.ASSOC_GATE = 2.0
self.initData()
class OtherStuffWidget(QWidget):
def __init__(self, parent=None):
super().__init__(parent=parent)
self.initUI()
self.d_m = DataManager()
self.d_m.add_event_handler(EventType.pose, self.newPoseData)
def initUI(self):
self.frame = QFrame(self)
self.frame.setStyleSheet('background-color: white;')
self.gestures = []
self.files = [
"images/make_fist.png", "images/wave_right.png",
"images/wave_left.png", "images/spread_fingers.png",
"images/unlock_gesture.png"
]
for i in range(5):
g = QLabel(self)
g.setPixmap(QPixmap(self.files[i]))
g.setScaledContents(True)
self.gestures.append(g)
def setGeometry(self, *__args):
super().setGeometry(__args[0], __args[1], __args[2], __args[3])
self.frame.resize(__args[2], __args[3])
space = 10
x = space
y = space
w = (__args[2] - space * 3) / 2
if w * 2 + space * 3 > __args[2]:
w = (__args[2] - space * 3) / 2
h = w
if h * 3 + space * 4 > __args[3]:
h = (__args[3] - space * 4) / 3
w = h
for i in range(5):
g = self.gestures[i]
g.setGeometry(x, y, w, h)
if (i + 1) % 2 == 0:
x = space
y = y + space + h
else:
x = x + space + w
g.setStyleSheet("border-radius: " + repr(int(w / 2)) +
"px; background-color: white;")
def newPoseData(self, event):
data = event["data"]
pose = int(data["pose"])
if pose == 0:
for g in self.gestures:
g.setStyleSheet(g.styleSheet().replace("yellow", "white"))
else:
g = self.gestures[pose - 1]
g.setStyleSheet(g.styleSheet().replace("white", "yellow"))
def evaluate(data_set, checkpoint_dir = 'tmp/train_data'):
with tf.Graph().as_default():
# Don't specify number of epochs in validation set, otherwise that limits the training duration as the
# validation set is 10 times smaller than the training set
#images, labels = read_data.inputs(data_set=data_set, batch_size=BATCH_SIZE, num_epochs=None)
if(data_set=="train"):
images, labels =DataManager.read_tfr_queue(train_cnn.DATA_SOURCE_TRAIN,BATCH_SIZE)
else:
images, labels = DataManager.read_tfr_queue(DataManager.TFR_SAVE_DIR + train_cnn.DATA_SOURCE_VALIDATION, BATCH_SIZE)
logits = model_cnn.inference(images)
accuracy_curr_batch = model_cnn.evaluation(logits, labels)
# Restore moving averages of the trained variables
mov_avg_obj = tf.train.ExponentialMovingAverage(model_cnn.MOVING_AVERAGE_DECAY)
variables_to_restore = mov_avg_obj.variables_to_restore()
saver = tf.train.Saver(variables_to_restore)
with tf.Session() as sess:
ckpt = tf.train.get_checkpoint_state(checkpoint_dir)
if ckpt and ckpt.model_checkpoint_path:
saver.restore(sess, ckpt.model_checkpoint_path)
else:
print('No checkpoint file found at %s' % checkpoint_dir)
return
coord = tf.train.Coordinator()
try:
threads = []
for qr in tf.get_collection(tf.GraphKeys.QUEUE_RUNNERS):
threads.extend(qr.create_threads(sess, coord, daemon=True, start=True))
num_iter = int(math.ceil(NUM_VALIDATION_EXAMPLES / BATCH_SIZE))
step = 0
acc_full_epoch = 0
while step < num_iter and not coord.should_stop():
acc_batch_val = sess.run(accuracy_curr_batch)
acc_full_epoch += acc_batch_val
step += 1
acc_full_epoch /= num_iter
tf.summary.scalar('validation_accuracy', acc_full_epoch)
summary_op = tf.summary.merge_all()
#summary_writer = tf.train.SummaryWriter(EVAL_DATA_DIR)
summary_writer = tf.summary.FileWriter(EVAL_DATA_DIR, sess.graph)
summary_str = sess.run(summary_op)
summary_writer.add_summary(summary_str, step)
print('Accuracy on full %s dataset = %.1f' % (data_set, acc_full_epoch))
except Exception as e:
coord.request_stop(e)
coord.request_stop()
coord.join(threads)
def LoadTestData(self, NumberOfImagesToLoad, TestWriteToFileFeature):
dm = DataManager.DataManager(True)
# Locate the test folder where we store the test images.
ParnetTestFolderPath = "C:\\FirebaseTestImages"
ChildTestFolderPath = "C:\\FirebaseTestImages\\Images"
# Retrieve all the contents from the test directory
parentFolderFilesList = os.listdir(ParnetTestFolderPath)
filename = parentFolderFilesList[0]
for a in range(NumberOfImagesToLoad):
# Rename the test image file and move it to the Images folder.
oldFileName = "{}\\{}".format(ParnetTestFolderPath, filename)
newFileName = "{}.png".format(self.GetCurrentDateTimeStamp())
newFileLocation = "{}\\{}".format(ChildTestFolderPath, newFileName)
fbNewFileLocation = "images/{}".format(newFileName)
#Copy the file to child folder, TestImages
shutil.copy(oldFileName, "{}\\{}".format(ChildTestFolderPath,
newFileName))
print("New File Copied: {}\\{}".format(ChildTestFolderPath,
newFileName))
# Add the new image to images folder in Firebase
storage = self.firebase.storage()
result = storage.child(fbNewFileLocation).put(newFileLocation)
# Get the url of the newly image added to Firebase
userToken = storage.credentials.access_token
firebaseImageURL = storage.child(fbNewFileLocation).get_url(
userToken)
# Retrieve latitude and longitude to add to our test data.
g = geocoder.google('Bluffton, SC')
latitude = g.latlng[0]
longitude = g.latlng[1]
# Add test data to the YardsTasks table
testData = {
"Latitude": latitude,
"Longitude": longitude,
"CreatedDate": dm.GenerateDateTimeStamp(),
"ModifiedDate": dm.GenerateDateTimeStamp(),
"UserID": self.testUserID,
"ImageURL": firebaseImageURL,
"ImageName": newFileName,
"Tags": "SOD",
"ImageProcessed": 0,
"TaskComplete": 0,
"ImageClassified": ''
}
db = self.firebase.database()
result = db.child("YardTasks").push(testData)
if (TestWriteToFileFeature):
dm = DataManager.DataManager(True)
dm.WriteTasksToFile(result['name'], newFileName)
time.sleep(3)
def onSuccessRequest(bodyData):
Logger.log("is onSuccessRequest:: ---- %s" % str(bodyData))
if bodyData != None and bodyData.get("list") != None:
listData = bodyData.get("list")
Logger.log("item.count = %s" % len(listData))
if (len(listData) <= 0):
return
# for classifyList in listData:
DataManager.saveJsonDataToFile(m_csvFileName, listData)
def __init__(self, parent, width=None, height=None):
OwnFrame.__init__(self, parent, width, height)
self.frame = super().getFrame()
self.dataManager = DataManager()
self._arrangeUI()
self._retranslateAll()
self._textInputRestrict()
self._arrangeDataInWidgets()
def main():
"""
"""
books = scrap()
for k in range(len(books)):
DataManager.managecsv(books[str(k)])
pass
def makeHeartbeatCall():
from startServer import app
with app.test_request_context():
try:
res = pyrequests.post(DataManager.getHeartbeatEndpoint(), headers=jsonHeaders,
data=DataManager.getHeartbeatJson())
if res.status_code != 200:
LedChanger.lightErrorLedOn()
except Exception as e:
LedChanger.lightErrorLedOn()
utils.printException(e)
def setup(self):
#print("in")
#print(self.grid)
try:
self.prologConnector = DataManager(self.row, self.column,
self.grid)
self.prologConnector.setup()
self.availableSlot = self.prologConnector.getLenPark()
except IndexError as e:
message = "The file you are trying to load has an invalid dimension.\n Valid Dimensions is 25x25"
messagebox.showerror("Invalid Dimension", message)
m = Menu()
m.main_loop()
def start_Process():
user_reply = input(colored.yellow('Save data to file? (y/n): '))
if (user_reply == 'yes') or (user_reply == 'y'):
keep_time = input(colored.yellow('Set Time Limit for Data Collection?(y/n): '))
if keep_time == 'yes' or keep_time == 'Yes' or keep_time == 'y':
DataManager.write_tofile( 0, True, True)
else:
size = input(colored.yellow('File Size(number of lines): '))
DataManager.write_tofile(int(size), False, True)
elif user_reply == 'n' or user_reply == 'no':
keep_time = input(colored.yellow('Set Time Limit for Data Collection?(y/n): '))
if keep_time == 'yes' or keep_time == 'Yes' or keep_time == 'y':
DataManager.write_tofile( 0, True, False)
else:
size = input(colored.yellow('File Size(number of lines): '))
DataManager.write_tofile(int(size), False, False)
else:
puts(colored.red('---Invalid Input---'))
def execute(self, context):
self.report({'INFO'}, "Training model %s..." %
context.scene.speech2anim_data.training_videos_path)
os.chdir(
bpy.utils.user_resource("SCRIPTS", "addons") + config.ADDON_PATH +
'/src')
reloadConfig(context)
#Clean previous output
#pdb.set_trace()
wops.rmdir(
config.TEMPDATA_OUTPUT[3:],
wops.clear(
bpy.utils.user_resource("SCRIPTS", "addons") +
config.ADDON_PATH))
wops.mkdir(
config.TEMPDATA_OUTPUT[3:],
wops.clear(
bpy.utils.user_resource("SCRIPTS", "addons") +
config.ADDON_PATH))
d = context.scene.speech2anim_data
DataManager.Train(d.training_videos_path, d.training_model_path)
paths = DataManager.getTrainingVideoPaths(d.training_videos_path)
#TODO: refactor
#for every video in the training folder
for path in paths:
#get the name
name = path.split('/')[-1:][0].split('.')[0]
exists = False
#if we don't have it in the list, add it
for p in d.training_videos_list:
if p.name == name:
exists = True
if not exists:
item = d.training_videos_list.add()
item.path = path
item.name = name
for i, p in enumerate(d.training_videos_list):
exists = False
for path in paths:
name = path.split('/')[-1:][0].split('.')[0]
if p.name == name:
exists = True
if not exists:
d.training_videos_list.remove(i)
return {'FINISHED'}
def thread_handle_sending_and_deleting_image(imagePath):
try:
with open(imagePath, 'rb') as img:
imageBasename = os.path.basename(imagePath)
files = {'img': (imageBasename, img, 'multipart/form-data')}
with pyrequests.Session() as s:
print("DataSender, starting to post image to server: " + str(imagePath))
r = s.post(DataManager.getPhotoReceiveEndpoint(), files=files)
print("DataManager.getPhotoReceiveEndpoint(), status code: " + str(r.status_code))
if r.status_code == 200:
DataManager.deleteFile(imagePath)
except Exception as e:
utils.printException(e)
DataManager.makeStorageCheck()
def init(self, pDataPath):
self._dataManager = DataManager(pDataPath, self._threshold,
self._limit, self._system)
self._dataManager.readData()
self._trackers = []
self._printer = BoxPrinter(self._paletteWidth, self._thicknessLine)
self._idLKEnumerator = 1
self._idCSRTEnumerator = 1
self._printStack = {}
self._sortActivated = None
self._centroidActivated = None
def PrepareData(name, cut):
p, pbar, weights, weightsbar = dm.ReadRealData(
name, cut
) # get the particle dictionaries and weights, splits regular to conjugate
p = BoostIntoRest(p) # Boosts particles into the COM frame
pbar = BoostIntoRest(pbar)
return p, pbar, weights, weightsbar
def GetData():
X = FM.FeaturesData()
y = DM.CreateSalesFrame()
for i in y.index:
if y['Week Number'][i] < 36 and y['Year'][i] == 2012:
y = y.drop([i])
elif y['Week Number'][i] > 44 and y['Year'][i] == 2019:
y = y.drop([i])
y = y.drop(columns=['Year', 'Week Number'])
#We create training and testing data that fit with sklearn package
X_train, X_test, y_train, y_test = ms.train_test_split(X,
y,
test_size=0.20,
random_state=0,
shuffle=False)
#Do we need this ?
sc_X = StandardScaler()
X_train = sc_X.fit_transform(X_train)
X_test = sc_X.transform(X_test)
return X_train, X_test, y_train, y_test
def test3(self):
self.dataManagerTest = DM.DataManager(
self.params['ModelParams']['dirTest'],
self.params['ModelParams']['dirResult3'],
self.params['DataManagerParams'])
self.dataManagerTest.loadTestData()
net = caffe.Net(
self.params['ModelParams']['prototxtTest'],
os.path.join(
self.params['ModelParams']['dirSnapshotsONOFF'],
"_iter_" + str(self.params['ModelParams']['snapshotONOFF']) +
".caffemodel"), caffe.TEST)
numpyImages = self.dataManagerTest.getNumpyImages()
# originNumpy = self.dataManagerTest.getNumpyImages()
for key in numpyImages:
mean = np.mean(numpyImages[key][numpyImages[key] > 0])
std = np.std(numpyImages[key][numpyImages[key] > 0])
numpyImages[key] -= mean
numpyImages[key] /= std
results = dict()
for key in numpyImages:
btch = np.reshape(numpyImages[key], [
1, 1, numpyImages[key].shape[0], numpyImages[key].shape[1],
numpyImages[key].shape[2]
])
net.blobs['data'].data[...] = btch
print numpyImages[key].shape
out = net.forward()
l = out["labelmap"]
print l.shape
labelmap = np.squeeze(l[0, 1, :, :, :])
res = np.squeeze(labelmap)
res = np.transpose(res, [2, 1, 0])
res = np.transpose(res, [1, 0, 2])
w = sitk.ImageFileWriter()
filename, ext = splitext(key)
# io.imsave("/home/quan/Desktop/VNet/Results/" + filename + "_rotate" + ext,np.squeeze(res))
# w.Execute(labelmap)
# results[key] = np.squeeze(labelmap)
# print "write result"
# toWrite = sitk.GetImageFromArray(results[key],isVector=False)
# toWrite = sitk.Cast(toWrite, sitk.sitkUInt8)
writer = sitk.ImageFileWriter()
# filename, ext = splitext(key)
writer.SetFileName(
os.path.join(self.params['ModelParams']['dirResult3'],
filename + "_rotate" + ext))
# writer.Execute(toWrite)
# original = np.squeeze(oribatch[0,0,:,:,:])
im2 = sitk.GetImageFromArray(np.squeeze(res), isVector=False)
im2 = sitk.Cast(sitk.RescaleIntensity(im2), sitk.sitkUInt8)
# writer.SetFileName("/home/quan/Desktop/VNet/TrainResult/" + filename + "_original" + ext)
writer.Execute(im2)
def run(self):
dataManagerTest = DM.DataManager(
vnet_config.params['ModelParams']['dirTest'],
vnet_config.params['ModelParams']['dirResult'],
vnet_config.params['DataManagerParams'])
#are you serious @fausto?????
dataManagerTest.createImageFileList()
dataManagerTest.loadImages()
dataManagerTest.createGTFileList()
dataManagerTest.loadGT()
volumes = dataManagerTest.getNumpyImages()
labels = dataManagerTest.getNumpyGT()
#print dataManagerTest.sitkImages
#print dataManagerTest.sitkGT
#print volumes
#print labels
#inputs = [(1,2,3), (4,5,6), (6,7,8), (1,2,3)]
#yield (01, volumes['image01.nii'], labels['label01.nii'])
for key in volumes:
image_num = re.findall("\d+", key)[0]
print key
print image_num
yield (image_num, volumes[key],
labels['label' + str(image_num) + '.nii'])
def __init__(self, parent=None):
super().__init__(parent=parent)
self.data_manager = DataManager()
self.data_manager.myo.interval.connect(self.interval)
self.data_manager.add_event_handler(EventType.connected, self.connectionChanged)
self.data_manager.add_event_handler(EventType.disconnected, self.connectionChanged)
self.data_manager.add_event_handler(EventType.rssi, self.newRssiData)
self.data_manager.add_event_handler(EventType.battery_level, self.newBatteryData)
self.timer = QTimer()
self.timer.timeout.connect(self.updateBattery)
self.last_battery = -1
self.initUI()
def makeNgrokAddressesCall():
from startServer import app
with app.test_request_context():
try:
res = pyrequests.post(DataManager.getNgrokAddressesEndpoint(), headers=jsonHeaders,
data=DataManager.getNgrokAddressesAsJson(), timeout=10)
if res.status_code != 200:
if res.status_code == 404: # Hotfix!!
return True
LedChanger.lightErrorLedOn()
return False
return True
except Exception as e:
LedChanger.lightErrorLedOn()
utils.printException(e)
return False
def processScannedImages(self):
newImagesForTrain =[]
for image in self.image_processing_list:
pageNum = Check_image_page(image.imagePath)
newImagesForTrain= newImagesForTrain + ExportHandriteLinesFromScannedDoc(image, pageNum)
numS, numE = DataManager.Insert_to_database(newImagesForTrain)
return (numS, numE)
def MultiSampleDalitzParameters(particles, CP=False, splitNum=100):
data = dm.SplitEvents(particles,
splitNum) # splits events into smaller sets
parameters = []
progress = 0
"""Calcualte CM variables and C_T"""
for d in data:
progress += 1
print("\r" + str(round(progress / len(data) * 100, 2)), end="")
params = DalitzParameters(d,
CP) # calulate statistics for the data set
parameters.append(params)
new_list = []
"""Merges each CM variable and C_T calulated for each data set"""
for i in range(6):
subset = []
for j in range(len(parameters)):
subset.append(parameters[j][i])
new_list.append(subset)
final_data = []
"""Puts the calculated values in a single list"""
for i in range(6):
final_data.append(np.concatenate(new_list[i]))
return final_data
def MultiSampleDalitzParameters(particles, split=10):
data = dm.SplitEvents(particles, split) # splits events into smaller sets
parameters = []
progress = 0
"""Calcualte CM variables"""
for d in data:
progress += 1
print(progress/len(data) * 100)
params = DalitzParameters(d) # calulate statistics for the data set
parameters.append(params)
new_list = []
"""Merges each CM variable and C_T calulated for each data set"""
for i in range(5):
subset = []
for j in range(len(parameters)):
subset.append(parameters[j][i])
new_list.append(subset)
final_data = []
"""Puts the calculated values in a single list"""
for i in range(5):
subset = np.array(new_list[i]) # converts list into an array
subset = list(subset.flatten('F')) # flattens this list into a matrix, columns are differnet CM variables for different C_T
lower = subset[:int(len(subset)/2)] # gets C_T < 0 states
lower = np.concatenate(lower).ravel() # merge columns into one column
upper = subset[int(len(subset)/2):] # # gets C_T > 0 states
upper = np.concatenate(upper).ravel()
subset = [lower, upper] # creates a list of uppr and lower values for the single CM variable
final_data.append(subset)
return final_data
def test(self, snapnumber):
# produce the results of the testing data
torch.cuda.set_device(self.params['ModelParams']['device'])
self.dataManagerTesting = DMoriginal.DataManager(
self.params['ModelParams']['dirTest'],
self.params['ModelParams']['dirResult'],
self.params['DataManagerParams'],
self.params['TestParams']['ProbabilityMap'])
self.dataManagerTesting.loadTestData()
#model = resnet3D.resnet34(nll = False)
model = vnet.VNet(nll=False)
prefix_save = os.path.join(self.params['ModelParams']['dirSnapshots'],
self.params['ModelParams']['tailSnapshots'])
name = prefix_save + str(snapnumber) + '_' + "checkpoint.pth.tar"
checkpoint = torch.load(name)
# load the snapshot into the model
model.load_state_dict(checkpoint['state_dict'])
model.cuda()
#produce the segementation results
results = self.getTestResultImages(
model, self.params['TestParams']['ProbabilityMap'])
for key in results:
self.dataManagerTesting.writeResultsFromNumpyLabel(
results[key], key)
def __splitCurrentRecording(self):
try:
import DataManager
from DataManager import deviceName, videoDir
DataManager.createVideoDirIfNotExists()
_video_path = str(videoDir) + str(deviceName) + "_" + str(datetime.datetime.now()).replace(" ", "_") + '.h264'
print("c")
self.__camera.split_recording(_video_path)
print("d")
path_to_return = self.__previous_monitoring_video_path
self.__previous_monitoring_video_path = _video_path
return path_to_return
except Exception as e:
utils.printException(e)
return None
def leer_entrenamiento():
training, header = dm.read_csv()
os.system("cls")
print("Se leyeron", len(training), "datos.\n")
os.system("pause")
creador_de_arboles(training, header)
pass
def leer_test(tree):
testing, header = dm.read_csv()
os.system("cls")
print("Se leyeron", len(testing), "datos.\n")
os.system("pause")
menu_predicciones(tree, testing, header)
pass
def test(self):
self.dataManagerTest = DM.DataManager(self.params['ModelParams']['dirTest'], self.params['ModelParams']['dirResult'], self.params['DataManagerParams'])
self.dataManagerTest.loadTestData()
net = caffe.Net(self.params['ModelParams']['prototxtTest'],
os.path.join(self.params['ModelParams']['dirSnapshots'], "_iter_" + str(self.params['ModelParams']['snapshot']) + ".caffemodel"),
caffe.TEST)
numpyImages = self.dataManagerTest.getNumpyImages()
for key in numpyImages:
mean = np.mean(numpyImages[key][numpyImages[key] > 0])
std = np.std(numpyImages[key][numpyImages[key] > 0])
numpyImages[key] -= mean
numpyImages[key] /= std
results = dict()
for key in numpyImages:
btch = np.reshape(numpyImages[key], [1, 1, numpyImages[key].shape[0], numpyImages[key].shape[1], numpyImages[key].shape[2]])
net.blobs['data'].data[...] = btch
out = net.forward()
l = out["labelmap"]
labelmap = np.squeeze(l[0, 1, :, :, :])
results[key] = np.squeeze(labelmap)
self.dataManagerTest.writeResultsFromNumpyLabel(np.squeeze(labelmap), key)
def update_context(self, action='APPEND'):
"""
Context Broker updateContext function
:param action: update context action ['APPEND', 'UPDATE', 'DELETE']
:rtype : requests.models.Response
"""
if action not in ['APPEND', 'UPDATE', 'DELETE']:
msg = "ContextBroker.update_context():The action passed to the function was not valid"
DM.data_manager_error(msg)
if len(self.entity.get_entity_list()) == 0:
msg = "ContextBroker.update_context(): Empty entity_list was passed to the function"
DM.data_manager_error(msg)
payload = {'contextElements': self.entity.get_entity_list(),
'updateAction': action}
data = json.dumps(payload)
url = self.CBurl+'/v1/updateContext'
response = self.get_response(data, url)
if response.status_code == 401:
msg = "ContextBroker.query_context(): User token not authorized."
DM.data_manager_error(msg)
self.clean_all()
return response
def get_auth_token(self):
"""
Returns token IDM .
:rtype : unicode
"""
try:
file_path = os.path.realpath(os.path.abspath(os.path.split(inspect.getfile(inspect.currentframe()))[0]))
if not os.path.exists('%s/auth/auth.dat' % file_path):
if not os.path.exists('%s/auth' % file_path):
os.mkdir('%s/auth' % file_path)
with open('%s/auth/auth.dat' % file_path, 'w') as json_file:
j_data = json.dumps({'token': ''})
json_file.write(j_data)
json_file.close()
with open('%s/auth/auth.dat' % file_path, 'r') as json_file:
self.token = json.loads(json_file.read())['token']
json_file.close()
except Exception as e:
msg = "OrionAction.get_auth_token(): %s" % e
DM.data_manager_error(msg)
def get_response(self, data, url):
"""
Context Broker request
:param data:
:param url:
:rtype : requests.models.Response
"""
try:
if self.orion:
headers = {'Content-Type': 'application/json', "X-Auth-Token": self.token, 'Accept': 'application/json'}
else:
headers = {'Content-Type': 'application/json', 'Accept': 'application/json'}
if self.tenant != '':
headers['Fiware-Service'] = self.tenant
if self.service_path != '':
headers['Fiware-ServicePath'] = self.service_path
response = requests.post(url, headers=headers, data=data)
return response
except requests.RequestException as e:
msg = "ContextBroker.get_response(): %s" % e.message
DM.data_manager_error(msg)
from DataManager import *
from hmm.Model import *
from hmm.ModelParams import *
import numpy as np
if __name__ == '__main__' :
datafile = '8_DATA.PCA_normalized.filtered.sample_zscores.RD.txt'
#datafile = 'data'
outputfile = 'output'
paramsfile = 'params.txt'
print 'Loading data file...'
dataloader = DataManager(datafile)
params = dataloader.getParams(paramsfile)
dataloader.skipHeadline()
sample = dataloader.getNextSample()
targets_list = dataloader.getTargetsList()
output = file(outputfile, 'w')
while sample :
#target_index is used to split observations sequence
target_index_begin = 0
target_index_end = 0
temp = 1
for targets in targets_list:
print 'Running HMM for sample[' + sample['sample_id'] + ']: ',
print 'chr' + targets[0]._chr + ' [' + str(temp) + '\\' + str(len(targets_list)) + ']'
temp += 1
target_index_end = target_index_begin + len(targets)
modelParams = ModelParams(params, targets)
#the 'observations' of sample is splitted
def discover(args) :
datafile = args.datafile
outputfile = args.output
paramsfile = args.params
sample_req = args.sample
sample_flag = False #used to check whether sample_req exists
print 'Loading data file...'
dataloader = DataManager(datafile)
params = dataloader.getParams(paramsfile)
dataloader.skipHeadline()
sample = dataloader.getNextSample()
targets_list = dataloader.getTargetsList()
output = file(outputfile, 'w')
output.write('SAMPLE_ID\tCNV\tFULL_INTERVAL\tINDEX\tINTERVAL\tREAD_DEPTH\n')
while sample :
if sample_req == '' or (sample_req != '' and sample['sample_id'] == sample_req):
sample_flag = True
#target_index is used to split observations sequence
target_index_begin = 0
target_index_end = 0
temp = 1
for targets in targets_list:
print 'Running HMM for sample[' + sample['sample_id'] + ']: ',
print 'chr' + targets[0]._chr + ' [' + str(temp) + '\\' + str(len(targets_list)) + ']'
temp += 1
target_index_end = target_index_begin + len(targets)
modelParams = ModelParams(params, targets)
#the 'observations' of sample is splitted
model = Model(modelParams, sample['observations'][target_index_begin:target_index_end])
pathlist = model.forwardBackward_Viterbi()
dataloader.outputCNV(output, sample['sample_id'], targets, pathlist, sample['observations'][target_index_begin:target_index_end])
target_index_begin = target_index_end
sample = dataloader.getNextSample()
output.close()
dataloader.closeFile()
if not sample_flag:
print 'Could not find the sample_id specified.'
def discover(args) :
paramsfile = args.params
sample_req = args.sample
hetsnp = args.hetsnp
tagsnp = args.tagsnp
vcf_file = args.vcf
if hetsnp == 'True' or hetsnp == 'TRUE':
hetsnp = True
else:
hetsnp = False
if tagsnp == 'True' or tagsnp == 'TRUE':
tagsnp = True
else:
tagsnp = False
datafile = args.rpkm_matrix
f_dir = os.path.dirname(datafile)
if f_dir != '':
f_dir = f_dir + '/'
if args.output:
outputfile = f_dir + str(args.output)
tagsnp_file = args.tagsnp_file
mode = args.mode
sample_flag = False #used to check whether sample_req exists
# Build a reference set
if mode == 'single' or mode == 'baseline' or mode == 'reference' or mode == 'ref':
print 'Building the reference dataset...'
dataloader = DataManager(datafile)
samples_np = dataloader.getAllSamples()
dataloader.closeFile()
print 'Baseline is Done.'
print 'Loading data file...',
dataloader = DataManager(datafile)
print 'Done!'
print 'Loading paramters...',
params = dataloader.getParams(paramsfile)
print 'Done!'
dataloader.skipHeadline()
sample = dataloader.getNextSample()
targets_list = dataloader.getTargetsList()
output_aux = file(outputfile+'.aux', 'w')
output_aux.write('SAMPLE_ID\tCNV_TYPE\tFULL_INTERVAL\tINDEX\tINTERVAL\tREAD_DEPTH\n')
output = file(outputfile,'w')
output.write('SAMPLE_ID\tCNV_TYPE\tINTERVAL\tCHROMOSOME\tSTART\tSTOP\tLENGTH\n')
if (hetsnp or tagsnp) and vcf_file == '':
print 'Error: please indicate a vcf file!'
system.exit(0)
if vcf_file != '':
vcf_reader = VCFReader(vcf_file)
else:
vcf_reader = False
if tagsnp:
print 'Loading tagSNP information ...',
cnp_dict = vcf_reader.loadTagSNP(tagsnp_file)
print 'Done!'
while sample :
if sample_req == '' or (sample_req != '' and sample['sample_id'] == sample_req):
sample_flag = True
print time.strftime("%Y-%m-%d %H:%M:%S", time.localtime()) ,sample_req,'......'
#Renjie added: To check whether the VCF contains sample_req.
vcf_checker = vcf.Reader(open(vcf_file,'r'))
if sample['sample_id'] in vcf_checker.samples:
sample_in_VCF = True
elif sample_req in vcf_checker.samples:
sample_in_VCF = True
else:
print 'No sample %s in VCF file.'%sample_req
sample_in_VCF = False
if hetsnp and sample_in_VCF :
print 'Parsing SNV information from VCF file for: ' + sample['sample_id']
snp_info = vcf_reader.getSNPInfo(sample['sample_id'], targets_list)
if tagsnp and sample_in_VCF:
print 'Analysing tagSNP information from tagSNP database for: ' + sample['sample_id'],
cnp_list = vcf_reader.findTagSNPForSample(sample['sample_pop'], sample['sample_id'], cnp_dict)
tagsnp_info_list = vcf_reader.findExonWithTagSNP(cnp_list, targets_list, overlap_threshold=0.5)
print len(tagsnp_info_list)
#estimate NB paramters from sample['observations']
sample_observations = []
remove_list = []
sample['observations'] = [ float(x) for x in sample['observations']]
#slicing: target_index is used to split observations sequence
target_index_begin = 0
target_index_end = 0
temp = 1
sample_observations_list = []
snp_info_list = []
for i, targets in enumerate(targets_list):
target_index_end = target_index_begin + len(targets)
if hetsnp and sample_in_VCF:
snp_info_list.append(snp_info[target_index_begin:target_index_end])
sample_observations_list.append(sample['observations'][target_index_begin:target_index_end])
target_index_begin = target_index_end
# Filtering:
if mode == 'svd' or mode == 'SVD' or mode == 'pooled' or mode == 'pooled-sample':
for i in range(len(sample_observations_list)):
sample_observations_list[i] = ndarray.tolist(stats.zscore(sample_observations_list[i]))
elif mode == 'baseline' or mode == 'reference' or mode == 'single' or mode == 'single-sample':
# filtering lists whose observation equals to 0
for i in range(len(targets_list)):
rem_index = []
for j in range(len(targets_list[i])):
value = sample_observations_list[i][j]
if np.isnan(float(value)):
rem_index.append(j)
#filter target_list, snp_list and observation_list
targets_list[i] = jf.filter_list_by_list(targets_list[i], rem_index)
sample_observations_list[i] = jf.filter_list_by_list(sample_observations_list[i], rem_index)
if hetsnp and sample_in_VCF:
snp_info_list[i] = jf.filter_list_by_list(snp_info_list[i], rem_index)
if tagsnp and sample_in_VCF:
tagsnp_info_list[i] = jf.filter_list_by_list(tagsnp_info_list[i], rem_index)
#Parameters estimation
observations_all_list = []
for i in range(len(sample_observations_list)):
observations_all_list.extend(sample_observations_list[i])
parameterLoader = ParameterEstimation(observations_all_list)
parameterList = parameterLoader.fit(observations_all_list,0.01,0.99)
print "Estimated Paramters: ",parameterList
params.append(parameterList[0])#mu
params.append(parameterList[1])#sd
for i, targets in enumerate(targets_list):
print 'Running HMM for sample[' + sample['sample_id'] + ']: ',
print 'chr' + targets[0]._chr + ' [' + str(temp) + '|' + str(len(targets_list)) + ']'
temp += 1
#Run the HMM
if not hetsnp and not tagsnp:
modelParams = ModelParams(mode, params, targets, het_nums=0, tagsnp=0)
elif sample_in_VCF and hetsnp and not tagsnp:
modelParams = ModelParams(mode, params, targets, snp_info_list[i], tagsnp=0)
elif sample_in_VCF and not hetsnp and tagsnp:
modelParams = ModelParams(mode, params, targets, het_nums=0, tagsnp=tagsnp_info_list[i])
elif sample_in_VCF and hetsnp and tagsnp:
modelParams = ModelParams(mode, params, targets, snp_info_list[i], tagsnp_info_list[i])
elif not sample_in_VCF and hetsnp and tagsnp:
modelParams = ModelParams(mode, params, targets, het_nums=0, tagsnp=0)
else:
pdb.set_trace()
model = Model(mode, modelParams, sample_observations_list[i])
pathlist = list()
if vcf_reader and sample_in_VCF:
pathlist = model.forwardBackward_Viterbi(mode, if_snp = True)
else:
pathlist = model.forwardBackward_Viterbi(mode, if_snp = False)
dataloader.outputCNVaux(output_aux, sample['sample_id'], targets, pathlist, sample_observations_list[i])
dataloader.outputCNV(output, sample['sample_id'], targets, pathlist, sample_observations_list[i])
sample = dataloader.getNextSample()
output.close()
output_aux.close()
dataloader.closeFile()
if not sample_flag:
print 'Could not find the sample_id specified.'
def game():
tick = 0
state = [[(list(), list()) for x in range(3)] for y in range(3)]
red_hq = HQ(TEAM.RED, 1, 0)
blue_hq = HQ(TEAM.BLUE, 1, 2)
state[0][1][TEAM.RED].append(red_hq)
state[2][1][TEAM.BLUE].append(blue_hq)
red_player_ai = RedPlayerAI(red_hq)
blue_player_ai = BluePlayerAI(blue_hq)
if VISUAL:
gs = pl.GridSpec(5, 1)
state_view = pl.subplot(gs[:3, :])
assets_view = pl.subplot(gs[3, :])
progress_view = pl.subplot(gs[4, :])
data_manager = DataManager()
data_manager.reset()
model = data_manager.get_model()
blue_player_ai.model = model
result = RESULT.DRAW
while True:
if red_player_ai.hq.hp <= 0 and blue_player_ai.hq.hp <= 0:
print "DRAW"
data_manager.add_win(RESULT.DRAW)
break
elif red_player_ai.hq.hp <= 0:
print "BLUE WIN"
result = RESULT.BLUE_WIN
data_manager.add_win(RESULT.BLUE_WIN)
break
elif blue_player_ai.hq.hp <= 0:
print "RED_WIN"
result = RESULT.RED_WIN
data_manager.add_win(RESULT.RED_WIN)
break
# print 'R', red_player_ai.money, red_hq.hp
# print 'B', blue_player_ai.money, blue_hq.hp
state = blue_player_ai.act(state, q=data_manager.transform(state, red_player_ai, blue_player_ai))
state = red_player_ai.act(state)
state = group_ai(state)
state = update(state)
red_player_ai.money += 1
blue_player_ai.money += 1
red_asstes, blue_assets = data_manager.evaluate_state(state, red_player_ai, blue_player_ai)
data_manager.add_sa(tick, red_player_ai, blue_player_ai, state, red_asstes, blue_assets)
state = update(state)
if VISUAL:
draw(state_view, assets_view, progress_view, state, data_manager)
tick += 1
# print
if VISUAL:
pl.pause(3)
data_manager.save()
return result
