Fixed Bug: KeyError: <ticker> for batch query

[Kuhno92]

For passing tickers as list for e.g. ['AAPL, 'GOOGL'], the concat operation results in KeyError: 'AAPL', because the df does not contain this index

[hydrosquall]

Hey @Kuhno92 , thanks for taking a look at this. Unfortunately, this fix breaks the current set of tests. Do you have a minimal code snippet that will reproduce the issue? The tests suggests that the current code should work when "tickers" is provided as a list of strings.

https://travis-ci.org/hydrosquall/tiingo-python/jobs/520429886

[Kuhno92]

Okay i have to admit that my suggested solution wasn't well thought out :)

I can reproduce the error with the following simple example:

from tiingo import TiingoClient

config = {}
config['session'] = True
config['api_key'] = "XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX"

client = TiingoClient(config)
ticker_history  = client.get_dataframe(['AAPL', 'GOOGL'], frequency='1min')

print(ticker_history)

This results in:

KeyError: 'AAPL'

I debugged this code, and it seems that the 'df.rename' function has no effect, because the variable 'metric_name' is none.

Due to this, indexing the df with the stock name fails.

As i understand this piece of code, the variable 'metric_name' is not set if the user does not specify a metric in the client.get_dataframe call.

[hydrosquall]

Hi @Kuhno92 ,

I took a closer look at this, and realized that we could do a better job of raising a more informative error message. The issue is that when passing in a list of tickers, you must provide a specific metric_name. This is intentional, but we could do a better job of providing an informative error message instead of a cryptic KeyError. I'll open a fix for this soon. In the meantime, this constraint is indicated in the docstring for the get_dataframe function.

https://github.com/hydrosquall/tiingo-python/blob/master/tiingo/api.py#L225-L226

Only single-ticker lookup supports single-metric lookup. If you'd like to get all the metrics back, for multiple tickers, here is something that might work for your purposes.

histories = []
tickers = ['GOOGL', 'APPL']
for ticker in tickers:
  history = client.get_dataframe(ticker, frequency='weekly')
  histories.append(history)

Let me know if that clears things up!

[hydrosquall]

@Kuhno92 , I've opened up a fix in #263 that will address this issue. Let me know if you have a chance to take a look / if you have any suggestions about what would make the error message more helpful. Thanks again for taking the time to take a look at this project and contribute the time for investigating!

[hydrosquall]

Replaced by #263, which changes the error message instead of changing behavior. The new feature request will instead be tracked under #265.