[Errno 22] Invalid argument

[dgarciagud]

Trying to obtain intraday data I obtain the following error:

data = hub.candle('AMZN', resolution='60', end="2020-05-22")
Traceback (most recent call last):

File "", line 1, in
data = hub.candle('AMZN', resolution='60', end="2020-05-22")

File "C:\Users\casa\Anaconda3\lib\site-packages\fhub\utils.py", line 143, in helper
return func(clase, *args, **kwargs)

File "C:\Users\casa\Anaconda3\lib\site-packages\fhub\utils.py", line 134, in helper
return func(clase, *args, **kwargs)

File "C:\Users\casa\Anaconda3\lib\site-packages\fhub\core.py", line 566, in candle
'from': _unixtime(start),

File "C:\Users\casa\Anaconda3\lib\site-packages\fhub\utils.py", line 169, in _unixtime
return int(datetime.strptime(_normalize_date(date), "%Y-%m-%d").timestamp())

OSError: [Errno 22] Invalid argument

[nicola-pesavento]

Hi, you should specify the start argument value when you use candle().

i.e. data = hub.candle('AMZN', resolution='60', start="2020-05-15", end="2020-05-22")

[arogima]

Apparently it's a Python error in Windows. I didn't find that bug when developing the code on Linux.

Timestamp conversion on windows fails with timestamps close to EPOCH

I'll find a way around it.

[nicola-pesavento]

@arogima is right, Windows has some problem to handle timestamp regarding dates before 1970 (see here). In particular, with my laptop (Windows 10) I have the same issue with all date before 1970/01/03 (yes 3rd January, I don't know why for the moment). I have just sent a pull request with a simple fix: if start argument is None then its value will be "1970-01-03".

[dgarciagud]

start="2020-05-15",

thanks it works when start date is introduced!

[paduel]

I update a new version 0.0.13, please @dgarciagud check if this version solved the bug in Windows.

[dgarciagud]

this is solved @paduel.
Additionally I have to deal with to minor errors

ImportError: cannot import name 'json_normalize' from 'pandas' (C:\Users\casa\Anaconda3\lib\site-packages\pandas_init_.py)

using from pandas.io.json import json_normalize instead of directly from pandas.

And another with NA from pandas.

[paduel]

this is solved @paduel.
Additionally I have to deal with to minor errors

ImportError: cannot import name 'json_normalize' from 'pandas' (C:\Users\casa\Anaconda3\lib\site-packages\pandas__init__.py)

using from pandas.io.json import json_normalize instead of directly from pandas.

And another with NA from pandas.

Nice!

You must to update Pandas for the other error.