for jhta.CRSI, could you please give a sample case. about jhtalib

[xiaods]

    # Get Binance Data into dataframe
    candles = client.get_klines(symbol='BTCUSDT', interval=Client.KLINE_INTERVAL_1MINUTE)
    df = pd.DataFrame(candles)
    df.columns=['timestart','open','high','low','close','?','timeend','?','?','?','?','?']
    df.timestart = [datetime.datetime.fromtimestamp(i/1000) for i in df.timestart.values]
    df.timeend = [datetime.datetime.fromtimestamp(i/1000) for i in df.timeend.values]

    # Compute RSI after fixing data
    float_data = [float(x) for x in df.close.values]
    np_float_data = np.array(float_data)
    rsi = talib.RSI(np_float_data, 14)
    df['rsi'] = rsi

     pp (jhta.CRSI(df.rsi, df.rsi, 3))

[joosthoeks]

You need two different DataFeeds. You can read the source here: https://www.fmlabs.com/reference/default.htm?url=RSIC.htm

basic usage:

pp (jhta.CRSI(df1, df2, 3, 'Close', 'Close'))

[xiaods]

@joosthoeks thanks for your clarify. got it.