Can't make candle chart due to some error with mpf.plot

Question

import pandas as pd
import yfinance as yf
import mplfinance as mpf

df = yf.download('AMZN', start='2020-01-01', end='2025-07-31')
print(df)
mpf.plot(df['2020-01-01':'2020-06-01'], type='candle', volume=True)

This is the code I have written.

This is the output:

Price            Close        High         Low        Open    Volume
Ticker            AMZN        AMZN        AMZN        AMZN      AMZN
Date
2020-01-02   94.900497   94.900497   93.207497   93.750000  80580000
2020-01-03   93.748497   94.309998   93.224998   93.224998  75288000
2020-01-06   95.143997   95.184502   93.000000   93.000000  81236000
2020-01-07   95.343002   95.694504   94.601997   95.224998  80898000
2020-01-08   94.598503   95.550003   94.321999   94.902000  70160000
...                ...         ...         ...         ...       ...
2025-07-14  225.690002  226.660004  224.240005  225.070007  35702600
2025-07-15  226.350006  227.270004  225.460007  226.199997  34907300
2025-07-16  223.190002  226.100006  222.179993  225.880005  39535900
2025-07-17  223.880005  224.500000  222.509995  223.320007  31855800
2025-07-18  226.130005  226.399994  222.979996  225.139999  37811600

[1393 rows x 5 columns]

This is the error:

    mpf.plot(df['2020-01-01':'2020-06-01'], type='candle', volume=True)
    ~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    dates,opens,highs,lows,closes,volumes = _check_and_prepare_data(data, config)
                                            ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^
raise ValueError('Column "'+col+'" NOT FOUND in Input DataFrame!'+
                     '\n            CHECK that your column names are correct AND/OR'+
                     '\n            CHECK for leading or trailing blanks in your column names.')
ValueError: Column "Open" NOT FOUND in Input DataFrame!
            CHECK that your column names are correct AND/OR

I tried to make candle stick chart using mplfinance but I could not. It seems that there is some error with the way I use the plot() function but I tried checking the syntax and as per my observations, it is correct.

Answer 1

It appears that, on or about Oct 21, 2024, with version 0.2.46, yf.download started returning, by default, a MultiIndex for the columns.

You can fix this by setting multi_level_index=False in your call to yf.download. This should work:

import pandas as pd
import yfinance as yf
import mplfinance as mpf

df = yf.download('AMZN', start='2020-01-01', end='2025-07-31', multi_level_index=False)
print(df)
mpf.plot(df['2020-01-01':'2020-06-01'], type='candle', volume=True)

---

How could you have found this?

1. Note carefully the error message:

ValueError: Column "Open" NOT FOUND in Input DataFrame!
CHECK that your column names are correct AND/OR
CHECK for leading or trailing blanks in your column names.

2. To check your column names:

print("df.columns =",df.columns)

3. The output would be:

df.columns= MultiIndex([( 'Close', 'AMZN'),
            (  'High', 'AMZN'),
            (   'Low', 'AMZN'),
            (  'Open', 'AMZN'),
            ('Volume', 'AMZN')],
           names=['Price', 'Ticker'])

which clearly shows what the error message is saying, Column "Open" NOT FOUND in Input DataFrame!

hth

---

I suspect that the reason yf.download() defaulted to MultiIndex is because you can, in theory, ask for multiple Tickers at the same time, and a MultiIndex will keep them separate. Personally I would prefer it to count the number of tickers and not use a MultiIndex unless I ask for more than one ticker. However possibly the authors did not want yf.download to vary it's return type based on the input (certainly a valid argument) and so they chose to always return MultiIndex.

By the way, on a separate note: I am also seeing the following message from yf when I run it: YF.download() has changed argument auto_adjust default to True so if you don't want auto_adjust you may want to also set that to False. Your choice.

Full Disclosure: I am the maintainer of the mplfinance package.