Volatility stop problem

[aiv1]

Hi!
I get data from IBKR tws api and want to calculate the volatility stop. Results are strange. See last column:

      dateTime      open      high       low     close   volume        rsi        volStp
0   2022-12-19  11656.25  11656.25  11656.25  11656.25        0        NaN           NaN
1   2022-12-20  11639.00  11639.00  11639.00  11639.00        0        NaN           NaN
2   2022-12-21  11802.00  11802.00  11802.00  11802.00        0        NaN           NaN
3   2022-12-22  11512.00  11512.00  11512.00  11512.00        0        NaN           NaN
4   2022-12-23  11534.00  11534.00  11534.00  11534.00        0        NaN           NaN
..         ...       ...       ...       ...       ...      ...        ...           ...
266 2024-01-16  16916.50  17034.00  16855.00  16978.50  1026057  61.200523  2.942384e+06
267 2024-01-17  16825.00  16885.25  16689.00  16859.50   909154  55.975640  2.543360e+06
268 2024-01-18  16996.75  17126.00  16942.50  17117.00  1044730  63.280798  2.852702e+06
269 2024-01-19  17190.50  17472.00  17166.25  17463.25   981099  70.394639  2.466864e+06
270 2024-01-22  17528.25  17584.75  17434.25  17455.50   884567  70.067432  2.797342e+06

My code:

        df = pandas.DataFrame(self.data, columns=['dateTime', 'open', 'high', 'low', 'close', 'volume'])
        df['dateTime'] = pandas.to_datetime(df['dateTime'], format="%Y%m%d") 
        df.to_csv('MNQ_Hourly.csv')  

        df['rsi'] = talib.RSI(df['close'])

        quotes_list = [
            Quote(d,o,h,l,c,v) 
            for d,o,h,l,c,v 
            in zip(df['dateTime'], df['open'], df['high'], df['low'], df['close'], df['volume'] )
        ]
        results = indicators.get_volatility_stop(quotes_list)
        results_list = [r.sar for r in results]
        df['volStp'] = results_list
        print(df)

[DaveSkender]

For volatility stop, and many other indicators, they rely on an initialization window that determines how they'll start calculating. As a result, the response values for those incalculable periods will be None or NaN in your case. For this indicator specifically, that warmup period can vary, depending on how long it takes for an identifiable trend reversal.

The first trend will have None values since it is not accurate and based on an initial guess.

[aiv1]

same result with indicators.get_volatility_stop(quotes_list, 7, 3)

[DaveSkender]

Our unit tests are running okay. Are upper_band and lower_band similarly damaged?

[aiv1]

yes, same issue in bands

[LeeDongGeon1996]

@DaveSkender So, it's not the problem about None?
I'm wondering how the results will be like in the C# package.

[DaveSkender]

Looks fine in the C# package, v2.5.0 (same underlying engine in v1.2.0 of the Python package).
https://dotnetfiddle.net/ImEkl4

@aiv1 upload your raw quotes and I'll try to run them against the C# library. I'll also be able to see if there's something goofy happening with the source quote data that might be throwing off the results.

Any chance there's something happening in your DataFrames syntax?