samchaaa

(
    agg_flow
    .reset_index()
    .style
    # after this we are not working a a dataframe but a "styler" object
    .format({'cost': '${:,.2f}', 'datetime': '{:%Y/%m}/01',
            'percent_quarterly flow': '{:.1%}',
            'off_goal': '{:+.1%}',
            **{col: '{:.1f}' for col in ['cfs', 'total_flow', 'quarterly_flow']}},
            na_rep='Missing')

samchaaa

def get_max_min(prices, smoothing, window_range):
    # Get prices
    smooth_prices = prices['close'].rolling(window=smoothing).mean().dropna()
    # Get max and min for window
    local_max = argrelextrema(smooth_prices.values, np.greater)[0]
    local_min = argrelextrema(smooth_prices.values, np.less)[0]
    print('local_max, local_min', local_max, local_min)
    price_local_max_dt = []
    # iterate thru points
    for i in local_max:

samchaaa

# Concat all trades
trades_all = pd.DataFrame()
for x in results:
    trades = pd.DataFrame(x['trades'])
    trades['sym'] = x['ticker']
    trades_all = pd.concat([
        trades_all,
        trades
    ], axis=0)
trades_all = trades_all[trades_all['change'] != 0].dropna()

samchaaa

# Fade if price closes outside of running 1h % change (w/2h ema)... closes after 15m.
close_time = 1
pct_change = 4
smoothing = 8

results = []
for ticker in tqdm(sp):
    
    data = pd.read_json(json.load(open('./{}/{}.json'.format(data_path, ticker), 'r')))

samchaaa

def get_signals(close_time):
    
    """
    Creates signals for pybacktest.
    Note pybacktest specifically needs series named "buy", "sell", "short", "cover" to work.
    Check out tutorial notebook: https://nbviewer.jupyter.org/github/ematvey/pybacktest/blob/master/examples/tutorial.ipynb
    
    """

    bs = ohlc['C'] > ohlc['hb'] # sell signal

samchaaa

def hb_lb(ohlc, n, smoothing):
    
    """
    Creates a "high band" and "low band", based on the absolute pct movement for a given period. 
    This will trigger the mean reversion trades.
    
    'pct': percent change in the underlying, based on period n
    'abs_vol': the absolute exponential moving average of the pct movement
    
    """

samchaaa

if data_path not in os.listdir():
    os.mkdir(data_path)

# save as json
for ticker in tqdm(sp):    
    data = get_data(ticker, '2020-08-01', '2020-12-15')
    with open('./{}/{}.json'.format(data_path, ticker), 'w') as j:
        json.dump(data.to_json(), open('./{}/{}.json'.format(data_path, ticker), 'w'))

samchaaa

def get_data(symbol, start, end):
    results = pd.DataFrame()
    while True:
        if start < end:
            # 15m bars
            d = api.polygon.historic_agg_v2(symbol, 15, 'minute', _from=start, to=end, unadjusted=False, limit=None).df
            time.sleep(.2)
            start = d.index[-1].strftime('%Y-%m-%d')
            results = pd.concat([results, d], axis=0)
        else:

samchaaa

Lecture 1: Introduction to Research — [📝Lecture Notebooks] [▶️Video]
Lecture 2: Introduction to Python — [📝Lecture Notebooks] [▶️Video]
Lecture 3: Introduction to NumPy — [📝Lecture Notebooks] [▶️Video]
Lecture 4: Introduction to pandas — [📝Lecture Notebooks] [▶️Video]
Lecture 5: Plotting Data — [📝Lecture Notebooks] [[▶️Vide

samchaaa

def lookup(vin):
    m = getManu(vin)
    try:
        y = int(getYear(vin))
    except Exception as e:
        y = e
    # Exceptions here, or Sam may clean up in dict
    # if m!= None:
    #    if 'General Motors' in m:
    #        print(m)