No bundle registered with the name 'sep'

octaviodegodoy commented 4 years ago

Hi, I was following the instructions for installing on window, but after running the script load_quandl_sf1.py I got the error "No bundle registered with the name 'sep' "

Updated with zipline ingest, set my quandl key properly but could not solve it. Thank you for any help.

FrankPSch commented 4 years ago

Hi Luiz,

I am currently also working on a proper installation and have been able to solve your questions yesterday ;).

Some questions to you, my current trial is in brackets:

Which versions do you use for python (py36 from python.org), zipline (zipline-live2), broker connectivity (IBpy2 from zipline-live2)?

Some hints from me:

In the /data/loaders/ directory, you can find the file 'sep_quandl.py' which you need to run before the 'load_quandl_sf1.py'.
Inside this file, you will also find an explanation on how to register 'sep' in '~/.zipline/extension.py' which needs to be done before.
As I have not yet managed to run 'NASDAQ_sector_code_loader.py' or 'SHARADAR_sector_code_loader.py', I have commented out the line 'from alphacompiler.data.NASDAQ import NASDAQSectorCodes' in the examples.

Have fun, Frank

FrankPSch commented 4 years ago

Also:

In file 'sparse_data.py', the class SparseDataFactor requires that the path 'self.data_path' is correctly set.

carstenf commented 3 years ago

Zipline 1.4 requires more information for exchange

Following changes in sep_quandl.py are necessary for zipline 1.4

Metadate sep need to be changed to "NYSE", line 5

metadata_list.append((df_tkr.index[0], df_tkr.index[-1], df_tkr.index[-1] + pd.Timedelta(days=1), row0["ticker"], "NYSE", # zipline 1.4 change to NYSE row0["ticker"] # TODO: can we delete this? ) )

Than in line 110 change the following:

` exchange = {'exchange': 'NYSE', 'canonical_name': 'NYSE', 'country_code': 'US'} # changes for zipline 1.4
exchange_df = pd.DataFrame(exchange,index = [0]) # changes for zipline 1.4

    # write metadata
    #asset_db_writer.write(equities=pd.DataFrame(metadata_list,
    #                                            columns=METADATA_HEADERS))
    asset_db_writer.write(equities=pd.DataFrame(metadata_list,
                                                columns=METADATA_HEADERS),
                                                exchanges=exchange_df  # changes for zipline 1.4 
                                                )`

with this changes you should be able to read the registered bundle.

This is the base to fix "require argument country_code"

pbharrin commented 3 years ago

This is now fixed, make sure your loader sets the country code like this.

pbharrin / alpha-compiler

No bundle registered with the name 'sep' #9