def __init__(self, policy_params): self.ob_dim = policy_params['ob_dim'] self.ac_dim = policy_params['ac_dim'] self.weights = np.empty(0) # a filter for updating statistics of the observations and normalizing inputs to the policies self.observation_filter = get_filter(policy_params['ob_filter'], shape = (self.ob_dim,)) self.update_filter = True
def __init__(self, policy_params): self.ob_dim = policy_params['ob_dim'] self.ac_dim = policy_params['ac_dim'] # a filter for updating statistics of the observations and normalizing inputs to the policies self.observation_filter = get_filter(policy_params['ob_filter'], shape=(self.ob_dim, )) self.state_normalize = {"mean", None, "std", None} self.update_filter = True
def process_tracker(self,tracker): Filter = get_filter(tracker.filter) Loader = get_loader(tracker.loader) loader = Loader(tracker,Filter(self.interest_list,tracker.publisher)) media_list = loader.fetch() for media in media_list: if not media in self.submitted_list and \ not media in self.wait_list: media.fetch() self.wait_list.append(media)
def __init__(self, policy_params): self.ob_dim = policy_params['ob_dim'] self.ac_dim = policy_params['ac_dim'] self.weights = np.empty(0) # a filter for updating statistics of the observations and normalizing # inputs to the policies self.observation_filter = filter.get_filter( policy_params['ob_filter'], shape=(self.ob_dim,)) self.update_filter = True
def __init__(self, policy_params): self.ob_dim = policy_params['ob_dim'] self.ac_dim = policy_params['ac_dim'] self.weights = np.zeros((self.ac_dim, self.ob_dim), dtype=np.float64) #self.weights = np.zeros(self.ac_dim) # a filter for updating statistics of the observations and normalizing inputs to the policies self.observation_filter = get_filter(policy_params['ob_filter'], shape=(self.ob_dim,), mean=policy_params['initial_mean'], std=policy_params['initial_std']) #self.observation_filter = None self.update_filter = True
def __init__(self, agent_args, id_num=0): Base_ARS_Agent.__init__(self) self.ob_dim = agent_args['ob_dim'] self.ac_dim = agent_args['ac_dim'] self.weights = np.zeros((self.ac_dim, self.ob_dim), dtype=np.float64) # a filter for updating statistics of the observations and normalizing inputs to the policies self.observation_filter = get_filter(agent_args['ob_filter'], shape=(self.ob_dim, )) #################################### # auction stuff self.id = id_num self.active = True
def _set_filters(self, value): """Filters may be specified in a variety of different ways, including by giving their name; we need to make sure we resolve everything to an actual filter instance. """ if value is None: self._filters = () return if isinstance(value, basestring): filters = map(unicode.strip, unicode(value).split(',')) elif isinstance(value, (list, tuple)): filters = value else: filters = [value] self._filters = [get_filter(f) for f in filters]
def __init__(self, policy_params): self.ob_dim = policy_params['ob_dim'] self.ac_dim = policy_params['ac_dim'] self.policy = MlpPolicy('pi', policy_params['ob_space'], policy_params['ac_space'], hid_size=64, num_hid_layers=2) all_var_list = self.policy.get_trainable_variables() var_list = [ v for v in all_var_list if v.name.split("/")[1].startswith("pol") ] self.get_flat = U.GetFlat(var_list) self.set_from_flat = U.SetFromFlat(var_list) sess = tf.get_default_session() U.initialize() #print('sess',sess) #sess.run(tf.global_variables_initializer()) self.weights = self.get_flat() # a filter for updating statistics of the observations and normalizing inputs to the policies self.observation_filter = get_filter(policy_params['ob_filter'], shape=(self.ob_dim, )) self.update_filter = True
if __name__=='__main__': _stock = Stock(1301, 't', 100) entry = MyEntry() entry.stock = _stock p( entry.check_long_entry(0) ) p( entry.check_long_entry(1) ) print p(entry.check_short_entry(0)) p(entry.check_short_entry(1)) print my_exit = MyExit() my_exit.stock = _stock trade1 = entry.check_long(0) my_exit.check_exit(trade1,1) p( trade1.entry_price ) p( trade1.exit_price ) print stop = MyStop() trade3 = entry.check_long(0) print stop.get_stop(trade3, 0) trade4 = entry.check_short(1) print stop.get_stop(trade4, 1) print filter = MyFilter() print filter.get_filter(0) print filter.get_filter(1) print filter.get_filter(2) print filter.get_filter(3)
from flask import Flask, request, render_template from PIL import Image from io import BytesIO import base64 import urllib from filter import add_dog_filter, get_filter import numpy as np app = Flask(__name__) dog_filter = get_filter() @app.route('/') def index(): return render_template('index.html') @app.route('/submit',methods=["POST"]) def submit_file(): #img = request.files['imagefile'] img = Image.open(BytesIO(request.files['imagefile'].read())) imx = np.array(img) result = add_dog_filter(imx, dog_filter) res_img = Image.fromarray(result) inp_byte_io = BytesIO() img.save(inp_byte_io, 'PNG') inp_byte_io.seek(0)