def download_data_from_yahoo(symbol,
adjust=True,
backfill=False,
start_date=config.start_date,
end_date=config.end_date):
"""Generates a .csv file containing the high, low, open, close, volume, and adjusted close by date from Yahoo for the given symbol
For currency conversion rates, the format is like "USDCAD=X"
For the S&P500 index, Yahoo uses the ticker GSPX. This function is hardcoded to query Yahoo using GSPX instead of SP500, but save the data to SP500.
Parameters:
symbol : str
adjust : bool
Whether to incluse the AdjClose column or not. Uses my implementation, not Yahoo's
backfill : bool, optional
start_date : date, optional
end_date : date, optional
Returns:
dataframe
A dataframe containing high, low, open, close, volume, and adjusted close by date for the given symbol
"""
# Get some extra data to minimize refreshing
if start_date > config.start_date:
start_date = config.start_date
if end_date < config.end_date:
end_date = config.end_date
with utils.HiddenPrints():
df = yf.Ticker(symbol).history(auto_adjust=False,
start=start_date,
end=end_date)
# This causes an error with pandas, so we're using yfinance instead
# if dividends causes errors, go to the error and remove the 'eval'
# df = dr.get_data_yahoo(symbol, start_date, end_date)
# df = df.merge(dr.DataReader(symbol, 'yahoo-dividends', start_date, end_date), how='outer', left_index=True, right_index=True)
if symbol.upper() in yf.shared._ERRORS:
raise RemoteDataError("No data fetched for symbol " + symbol +
" using yfinance")
df.drop("Adj Close", axis=1, inplace=True)
df.drop("Stock Splits", axis=1, inplace=True)
if adjust:
df["AdjClose"] = df["Close"].add(df["Dividends"].cumsum())
if backfill:
utils.backfill(df)
# df.sort_index(inplace=True) # Yahoo data is sorted anyways
utils.debug(df)
df.to_csv(
utils.get_file_path(config.prices_data_path,
price_table_filename,
symbol=symbol))
return df
import os
sys.path.insert(0, '..')
import numpy as np
from gym.spaces import Discrete, Box
from featureExtractor.drone_feature_extractor import total_angle_between
from featureExtractor.drone_feature_extractor import DroneFeatureSAM1, DroneFeatureMinimal, DroneFeatureOccup, DroneFeatureRisk, DroneFeatureRisk_v2
from featureExtractor.drone_feature_extractor import DroneFeatureRisk_speed
from envs.drone_env_utils import InformationCollector
from alternateController.potential_field_controller import PotentialFieldController as PFController
from alternateController.social_forces_controller import SocialForcesController
from itertools import count
import utils # NOQA: E402
from envs.gridworld import GridWorld
from envs.drone_env_utils import angle_between
import copy
with utils.HiddenPrints():
import pygame
import pygame.freetype
def deg_to_rad(deg):
return deg*np.pi/180
def rad_to_deg(rad):
return rad*180/np.pi
def get_rot_matrix(theta):
'''
returns the rotation matrix given a theta value(radians)
'''
import os
import argparse
import utils as ut
desc_msg = 'MNIST segmentation using tensorflow and some sort of LeNet-5'
parser = argparse.ArgumentParser(desc_msg)
### START: Select the device where the operations will be executed
# All posible choices for device selection
# Hidde debug infro from tf... https://stackoverflow.com/a/38645250/5969548
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
from tensorflow.python.client import device_lib
with ut.HiddenPrints():
all_devices = device_lib.list_local_devices()
choices = []
for i in range(len(all_devices)):
choices.append(i)
device_default = choices[-1]
msg = 'Select which device tf should use for the ops.'\
+' 0: CPU, 1>=GPU (if available).'
parser.add_argument('-d','--device',help=msg,type=int,
choices=choices,default=device_default)