Klinger Volume Oscillator

[julian-dotcom]

Hi,

I am trying to use the Klinger Volume Oscillator found in pandas_ta.

To initialize this indicator, I try to use:

vbt.pandas_ta('kvo')

However, this always creates the following error message:

Traceback (most recent call last):
  File "/Users/julian/Documents/GitHub/bearbot/vectorbt_tests/klingerOscillator.py", line 53, in <module>
    klingerOscillator(symbol, start, interval)
  File "/Users/julian/Documents/GitHub/bearbot/vectorbt_tests/klingerOscillator.py", line 26, in klingerOscillator
    vbt.pandas_ta('kvo', index=2)
  File "/opt/homebrew/lib/python3.9/site-packages/vectorbt/indicators/__init__.py", line 30, in pandas_ta
    return IndicatorFactory.from_pandas_ta(*args, **kwargs)
  File "/opt/homebrew/lib/python3.9/site-packages/vectorbt/indicators/factory.py", line 3693, in from_pandas_ta
    config = cls.parse_pandas_ta_config(pandas_ta_func, **parse_kwargs)
  File "/opt/homebrew/lib/python3.9/site-packages/vectorbt/indicators/factory.py", line 3551, in parse_pandas_ta_config
    if not pd.Index.equals(result.index, test_df.index):
AttributeError: 'NoneType' object has no attribute 'index'

The same works perfectly well when I try to utilize the Chaikin Money Flow as follows:

CMF = vbt.pandas_ta('cmf')

Thanks in advance :)

[polakowo]

@julian-dotcom this is on pandas-ta's side, KVO returns None for whatever reason (https://github.com/twopirllc/pandas-ta/blob/bc3b292bf1cc1d5f2aba50bb750a75209d655b37/pandas_ta/volume/kvo.py#L22). I'll look into this.

[polakowo]

Try this: vbt.pandas_ta('kvo', parse_kwargs=dict(test_index_len=100)). The reason for this is that vectorbt needs to run the indicator on sample data to retrieve its outputs metadata, and by default, it constructs a test dataframe with 50 data points. Since one of the parameters of KVO is a rolling window of 55, pandas-ta returns None indicating that there is not enough data points. test_index_len lets you increase the number of rows in the test dataframe.

[julian-dotcom]

Wow, thank you for your effort! However, another bug arises next. When I update the code with your recommendation as follows:

    KVO = vbt.pandas_ta('kvo', parse_kwargs=dict(test_index_len=100))
    KVO.run(high=df['high'], low=df['low'], close=df['close'], volume=df['volume'])

I get the following error message:

Traceback (most recent call last):
  File "/Users/julian/Documents/GitHub/bearbot/vectorbt_tests/klingerOscillator.py", line 54, in <module>
    klingerOscillator(symbol, start, interval)
  File "/Users/julian/Documents/GitHub/bearbot/vectorbt_tests/klingerOscillator.py", line 27, in klingerOscillator
    KVO.run(high=df['high'], low=df['low'], close=df['close'], volume=df['volume'])
  File "/opt/homebrew/lib/python3.9/site-packages/vectorbt/indicators/factory.py", line 13, in run
    
  File "/opt/homebrew/lib/python3.9/site-packages/vectorbt/indicators/factory.py", line 2862, in _run
    results = run_pipeline(
  File "/opt/homebrew/lib/python3.9/site-packages/vectorbt/indicators/factory.py", line 1933, in run_pipeline
    output_list, other_list = _split_output(output)
  File "/opt/homebrew/lib/python3.9/site-packages/vectorbt/indicators/factory.py", line 1918, in _split_output
    raise ValueError("Number of returned outputs other than expected")
ValueError: Number of returned outputs other than expected

[polakowo]

@julian-dotcom as said, pandas-ta returns None whenever your data is not longer than your rolling windows. Just increase your data or reduce rolling windows.