After updating to 0.2.54, yf.download returns the price columns in different sort order

[dgarrard]

After updating to 0.2.54, yf.download returns the dataframe in this column order:
Date,Close,High,Low,Open,Volume
(which appears to be alphabetic for the price columns)

Previously, the dataframe was arranged in the more popular OHLCV format
Date,Open,High,Low,Close,Volume

Is there a setting that can be used to go back to OHLCV

[BiggRanger]

I'm assuming you're saving the data to a text file then reading it back in later, I'm doing the same. I parse the header of the file to ensure it matches what I expect.

I'm doing this for the check:

    # Note: standardHeader 1 and 2 are for legacy data imports from yfinance.history
    # standardHeader3 is for the data from yfinance.download which contains Close Adjusted and the correct Close data
    standardHeader1 = ['', 'Open', 'High', 'Low', 'Close', 'Volume', 'Dividends', 'Stock', 'Splits\n']
    standardHeader2 = ['', 'Open', 'High', 'Low', 'Close', 'Volume', 'Dividends', 'Stock', 'Splits',
                       'Capital', 'Gains\n']
    standardHeader3 = ['', 'Open', 'High', 'Low', 'Close', 'Adj', 'Close', 'Volume\n']
    standardHeader4 = ['Price','Adj','Close','Close','High','Low','Open','Volume\n']
    dataArray: list[PriceData] = []
    if os.path.exists(fileName):
        file = open(fileName, 'r')
        header = ''
        dateFormat = ''
        for index, text in enumerate(file):
            if text == 'Empty DataFrame\n':
                return []
            if index == 0:
                text = ShrinkWhitespace(text)
                header = text.split(' ')
                if (header != standardHeader1 and header != standardHeader2 and
                        header != standardHeader3 and header != standardHeader4):
                    raise Exception('!!!!! ERROR 0011 CHANGED HEADER IN PRICE FILE !!!!!')
           # rest of code to parse file here...



def ShrinkWhitespace(text: str):
    iterate = True
    while iterate:
        x = len(text)
        text = text.replace('  ', ' ')
        y = len(text)
        if x == y:
            iterate = False
    return text

[dgarrard]

Thanks @BiggRanger. I had done a similar thing in my code that reads the text file I create. Seems strange that the column sort order would change without warning or documentation on what is a very basic change.

[ValueRaider]

It is strange, not intended.

[optimvsjag]

I've the same problem, should be a bug. Now we have the 3 firsts rows with heads and the description not coincide with downloaded data.

[BiggRanger]

This changed in 0.2.51 I believe. I updated my parsers to handle the difference. Here is a write-up on the topic with yfinance and pandas: https://stackoverflow.com/questions/63107594/how-to-deal-with-multi-level-column-names-downloaded-with-yfinance/63107801#63107801

[dgarrard]

There is also discussion here: #2193