0

I have DF = { open: [], high:[], low:[], close[] )

and formula to aggregate current DF:

# OPEN = (Open_n-1 + Close_n-1) / 2.0
# HIGH = Max(High_n, Open_n, Close_n)
# LOW = Min(Low_n, Open_n, Close_n)
# CLOSE = (Open_n + High_n + Low_n + Close_n) / 4.0

Having issue with OPEN = (Open_n-1 + Close_n-1) / 2.0 part. No idea how to access previous row values... Maybe iterate through rows, generate prev row placeholder with another select query but it seems quite hacky.

Does anybody have suggestions to solve this one?

At the moment I have this:

self.candle_data_ctx.price_candle_df.
select([
    # OPEN = (Open_n-1 + Close_n-1) / 2.0
    #     TODO - implement
    # HIGH = Max(High_n, Open_n, Close_n)
    pl.max([pl.col('high_price'), pl.col('open_price'), pl.col('close_price')
                ]).alias('high_price'),
    # LOW = Min(Low_n, Open_n, Close_n)
    pl.min([pl.col('low_price'), pl.col('open_price'), pl.col('close_price')
                ]).alias('low_price'),
    # CLOSE = (Open_n + High_n + Low_n + Close_n) / 4.0
    (
       (
          pl.col('open_price') + pl.col('high_price') + pl.col('low_price') + pl.col('close_price')
        ) / 4.0
    ).alias('close_price'),
])
Joris Medeišis
  • 101
  • 4
  • You can `.shift()` to access previous/next rows - but this looks like it may be a form of https://stackoverflow.com/questions/75777384/how-to-generate-a-savings-simulation-with-python-and-polars-which-earns-interest - i.e. a cumulative calculation where the result is dependent on the result from the previous row - in which case you must iterate? – jqurious Mar 24 '23 at 15:55
  • Thanks, I will try .shift(). And -no the calculation is not cumulative, new OHLC values ar not connected. But thank you for the reference about cumulative too. – Joris Medeišis Mar 24 '23 at 16:15
  • Please do not add answers to the question body itself. Instead, you should add it as an answer. [Answering your own question is allowed and even encouraged](https://stackoverflow.com/help/self-answer). – Adriaan Mar 28 '23 at 08:34

1 Answers1

0

SOLVED - Heikin-Ashi candle calculation:

price_candle_df
.select([
    # OPEN_n = (Open_n-1 + Close_n-1) / 2.0
    (
        (
            pl.col('open_price').shift(periods=1) + pl.col('close_price').shift(periods=1)
        ) / 2.0
    ).fill_null(pl.col('open_price')).alias('open_price'),

    # HIGH_n = Max(High_n, Open_n, Close_n)
    pl.max([
        pl.col('high_price'),
        pl.col('open_price'),
        pl.col('close_price')
    ]).alias('high_price'),

    # LOW_n = Min(Low_n, Open_n, Close_n)
    pl.min([
        pl.col('low_price'),
        pl.col('open_price'),
        pl.col('close_price')
    ]).alias('low_price'),

    # CLOSE_n = (Open_n + High_n + Low_n + Close_n) / 4.0
    (
        (
            pl.col('open_price') + pl.col('high_price') + 
            pl.col('low_price') + pl.col('close_price')
        ) / 4.0
     ).alias('close_price')
])
Joris Medeišis
  • 101
  • 4