pandas-1.4 yfinance-0.1.69 AttributeError: 'Index' object has no attribute 'tz_localize'

[peterxie1]

Just upgraded to yfinance-0.1.69 and pandas-1.4 you will likely hit this issue. The code is trivial and used to work fine. Any suggestions?

python-3.9.10 windows 10

python Python 3.9.10 (tags/v3.9.10:f2f3f53, Jan 17 2022, 15:14:21) [MSC v.1929 64 bit (AMD64)] on win32 Type "help", "copyright", "credits" or "license" for more information.

import yfinance as yf yf.version '0.1.69' ticker = "QQQ" period = "1d" interval = "5m" print(ticker, period, interval) QQQ 1d 5m yf.download([ticker], period=period, interval=interval) Exception in thread Thread-3: Traceback (most recent call last): File "C:\Users\peter\AppData\Local\Programs\Python\Python39\lib\threading.py", line 973, in _bootstrap_inner self.run() File "C:\Users\peter\AppData\Local\Programs\Python\Python39\lib\threading.py", line 910, in run self._target(*self._args, **self.kwargs) File "C:\Users\peter\AppData\Local\Programs\Python\Python39\lib\site-packages\multitasking_init.py", line 104, in _run_via_pool return callee(*args, **kwargs) File "C:\Users\peter\AppData\Local\Programs\Python\Python39\lib\site-packages\yfinance\multi.py", line 188, in _download_one_threaded data = _download_one(ticker, start, end, auto_adjust, back_adjust, File "C:\Users\peter\AppData\Local\Programs\Python\Python39\lib\site-packages\yfinance\multi.py", line 202, in _download_one return Ticker(ticker).history(period=period, interval=interval, File "C:\Users\peter\AppData\Local\Programs\Python\Python39\lib\site-packages\yfinance\base.py", line 295, in history df.index = df.index.tz_localize("UTC").tz_convert( AttributeError: 'Index' object has no attribute 'tz_localize' Traceback (most recent call last): File "", line 1, in File "C:\Users\peter\AppData\Local\Programs\Python\Python39\lib\site-packages\yfinance\multi.py", line 112, in download _time.sleep(0.01) KeyboardInterrupt

[tsdeng]

lol, just saw the same problem and spent an hour debugging. It happens with pandas version 1.4 which is just released yesterday. To workaround it, you can install pip install pandas==1.3.5. But yeah, it's a problem that needs a fix when using pandas 1.4

[peterxie1]

Thanks for the note! yes, downgrading pandas from 1.4 to 1.3.5 worked-around this issue.

[wonka929]

yes, same issue here!

[nichmetsa]

same issue here

[v3natio]

same here, had to roll back the pandas version

[ysdede]

lol, just saw the same problem and spent an hour debugging. It happens with pandas version 1.4 which is just released yesterday. To workaround it, you can install pip install pandas==1.3.5. But yeah, it's a problem that needs a fix when using pandas 1.4

Thanks!

[Rational-IM]

I also tried downgrading yfinance to older versions - it didn't work. There were other changes on pandas 1.4 that required some adjustments on other parts of my code - maybe this update is a broader one.

[mundus08]

I faced the same problem. It might be related to pandas change: "Ignoring dtypes in concat with empty or all-NA columns" The problem seams to in df = _pd.concat([quotes, dividends, splits], axis=1, sort=True)

df.intext.dtype=object in pandas 1.4 because dividends and splits are empty. With pandas 1.3.5 df.intext.dtype=datetime64[ns]

[Dharmik-19]

Hi @mundus08 , I am a newbie in the deployment domain. I was curious how you find the exact lines which might be the solution to the problem we are facing. So, did you go through the entire code for pandas to find the prospective lines which can be causing the trouble?

[mundus08]

Hi @Dharmik-19 the stack trace contained the line number where the problem occured in yfinance. Then I added some print statements to the code and found out that the index dtype has changed. I suspected that it was due to the concat statement (because the dtype of the quotes df was still correct). In the release notes of pandas 1.4 I found a reference to the changes in the concat method.

[ProgrammingDao]

Duplicate of https://github.com/ranaroussi/yfinance/issues/937 I believe.

[Rational-IM]

Although #937 was closed, I still get an error message after updating Python/Pandas. I run the file test_yfinance.py and here are the results:

========================================= ERROR: test_attributes (main.TestTicker)

Traceback (most recent call last): File "C:\Users\danil\OneDrive\Python_projects\RIM_trading\untitled0.py", line 44, in test_attributes ticker.dividends File "C:\Users\danil\anaconda3\envs\spyder-5.2.2\lib\site-packages\yfinance\ticker.py", line 131, in dividends return self.get_dividends() File "C:\Users\danil\anaconda3\envs\spyder-5.2.2\lib\site-packages\yfinance\base.py", line 692, in get_dividends self.history(period="max", proxy=proxy) File "C:\Users\danil\anaconda3\envs\spyder-5.2.2\lib\site-packages\yfinance\base.py", line 295, in history df.index = df.index.tz_localize("UTC").tz_convert( AttributeError: 'Index' object has no attribute 'tz_localize'

========================================= ERROR: test_info_history (main.TestTicker)

Traceback (most recent call last): File "C:\Users\danil\OneDrive\Python_projects\RIM_trading\untitled0.py", line 35, in test_info_history history = ticker.history(period="max") File "C:\Users\danil\anaconda3\envs\spyder-5.2.2\lib\site-packages\yfinance\base.py", line 295, in history df.index = df.index.tz_localize("UTC").tz_convert( AttributeError: 'Index' object has no attribute 'tz_localize'

---

Ran 3 tests in 20.653s

FAILED (errors=2)

[neilsmurphy]

Confirm downgrading to pandas 1.3.5 solved the issue.

[grzesir]

any updates on a fix?

[hubbubba]

The pandas downgrade worked for me as well. I'm looking forward to a fix for 1.4x.

[ValueRaider]

I'm willing to fix but can't reproduce so I need your help to debug. I've created a branch with debugging code, can you use this (with latest Pandas obv): https://github.com/ValueRaider/yfinance/tree/fix/bad-index-dt

Don't mess with PIP just do a git fetch. Then run and paste the output:

import sys ; sys.path.insert(0, "path/to/my/branch")
import yfinance as yf
df = yf.download("QQQ", interval="5m", period="1d")

My output looks like this FYI:

Price table:
                           Open        High         Low       Close   Adj Close   Volume
2022-07-01 13:30:00  279.790009  280.250000  278.369995  278.619995  278.619995  3982900
2022-07-01 13:35:00  278.559998  278.679993  277.399994  278.029999  278.029999  1646519
2022-07-01 13:40:00  278.029999  280.000000  277.980011  278.859985  278.859985  1600816
2022-07-01 13:45:00  278.879913  281.079987  278.690002  280.550598  280.550598  1688162
2022-07-01 13:50:00  280.549988  281.559998  280.160004  281.209991  281.209991  1385621
...                         ...         ...         ...         ...         ...      ...
2022-07-01 19:40:00  282.165009  282.359985  281.910004  282.089996  282.089996   859425
2022-07-01 19:45:00  282.079987  282.380005  281.829987  282.290009  282.290009  1166711
2022-07-01 19:50:00  282.309998  282.529999  281.290009  281.809998  281.809998  1413182
2022-07-01 19:55:00  281.809998  282.279999  281.660004  282.119995  282.119995  2467601
2022-07-01 20:00:00  282.130005  282.130005  282.130005  282.130005  282.130005        0

[79 rows x 6 columns]
Index type = <class 'pandas.core.indexes.datetimes.DatetimeIndex'>
Index[0] = 2022-07-01 13:30:00
- type = <class 'pandas._libs.tslibs.timestamps.Timestamp'>
 - tz = None
 - tz type = {} <class 'NoneType'>
[*********************100%***********************]  1 of 1 completed

[ValueRaider]

Does this still happen in the latest release?

[waqaskhan137]

For me it was still causing the issue even after install panadas 1.3.5 But then I printed the version and came to know it wasnt using pandas 1.3.5 I had to restart the kernal of jupyter notebook in order to get the defined version. Hope this will help someone who is facing the similar issue.

!pip install yfinance
!pip install -U pandas==1.3.5

import yfinance as yf
import pandas as pd
import sys
 
 
print("Python Version", sys.version)
print("Yahoo Fianance lib : ", yf.__version__)
print("Panadas : ", pd.__version__)

Python Version 3.10.6 (main, Aug 10 2022, 11:40:04) [GCC 11.3.0] Yahoo Finance : 0.1.67 Panadas : 1.5.1

[wgong]

Thank you, pandas==1.3.5 worked

[ValueRaider]

@wgong Can you provide code that was failing before you downgraded Pandas? We want to fix this problem but we can't reproduce.

[wgong]

@ValueRaider I just upgraded pandas back to 1.5.1, and cannot reproduced it either.

Here is my function which failed before

def gwg_download_yf(tickers, start_dt, end_dt=""):
    """download quotes of multiple tickers for given start_dt, end_dt (exclusive)
    Args:
        tickers (list)
        start_dt (str) - YYYYMMDD
        end_dt (str) - YYYYMMDD, optional. If end_dt = "", will download one date at start_dt
        
    Returns:
        dataframe or None
    """    
    df_ = None
    if len(tickers) < 1:
        return df_
    
    start_date = datetime.strptime(start_dt, "%Y%m%d")
    if not end_dt:
        end_date = start_date + timedelta(days=1)
    else:
        end_date = datetime.strptime(end_dt, "%Y%m%d")
    data = yf.download(" ".join(tickers), 
                    start=datetime.strftime(start_date, "%Y-%m-%d"), 
                    end=datetime.strftime(end_date, "%Y-%m-%d"))
    return data

[wgong]

@ValueRaider I tried on a different Ubuntu computer, and cannot reproduce the initial error.

[AminatAzeez]

I downgraded to 1.3.5 and restarted the kernel but I am still getting the same error

AttributeError Traceback (most recent call last) Input In [10], in <cell line: 3>() 1 #Question 3 Obtain stock data for AMD using the history function, set the period to max. 2 #Find the Volume traded on the first day (first row). ----> 3 amd_stock_data = amd.history(period="max")

File ~\anaconda3\lib\site-packages\yfinance\base.py:279, in TickerBase.history(self, period, interval, start, end, prepost, actions, auto_adjust, back_adjust, proxy, rounding, tz, timeout, **kwargs) 276 df["Stock Splits"].fillna(0, inplace=True) 278 # index eod/intraday --> 279 df.index = df.index.tz_localize("UTC").tz_convert( 280 data["chart"]["result"][0]["meta"]["exchangeTimezoneName"]) 282 if params["interval"][-1] == "m": 283 df.index.name = "Datetime"

AttributeError: 'Index' object has no attribute 'tz_localize'

please help....

[ValueRaider]

That looks like old code. Update yfinance to latest version, confirm with import yfinance as yf ; print(yf.__version__)

[AminatAzeez]

I checked version and it says 0.1.67

[AminatAzeez]

It works fine now, I just realized the version of the pandas was not changing , I was installing it the wrong way. So, I got the correct installation and it worked

[iamNeverwell]

Restarted Kernel;

!pip install yfinance !pip install -U pandas==1.3.5

Confirmed

[ValueRaider]

Confirmed

Confirmed what?

Also everyone should be able to use any version of Pandas >=1.3.0 without problems.

[Abdul675]

In my case downgrading the pandas version to 1.3.5 is not working same error how can i resolve this

[waqaskhan137]

For me it was still causing the issue even after install panadas 1.3.5 But then I printed the version and came to know it wasnt using pandas 1.3.5 I had to restart the kernal of jupyter notebook in order to get the defined version. Hope this will help someone who is facing the similar issue.

!pip install yfinance
!pip install -U pandas==1.3.5

import yfinance as yf
import pandas as pd
import sys
 
 
print("Python Version", sys.version)
print("Yahoo Fianance lib : ", yf.__version__)
print("Panadas : ", pd.__version__)

Python Version 3.10.6 (main, Aug 10 2022, 11:40:04) [GCC 11.3.0] Yahoo Finance : 0.1.67 Panadas : 1.5.1

@Abdul675 please refer to this

[ValueRaider]

@Abdul675 Try branch fix/prices-index-not-datetime (instructions #1080) and give feedback.