Bug in Alphalens?

hannes · May 2, 2022, 3:52pm

Hi Brian,

According to this;

quantrocket-llc/alphalens/blob/1a6cb65f4844759e8ebb703847b24cf4a9ae5908/alphalens/performance.py#L933

    
      
              if quantiles is not None:
                  portfolio_data = portfolio_data[portfolio_data['factor_quantile'].isin(
                      quantiles)]
          
          
    if groups is not None:
                  portfolio_data = portfolio_data[portfolio_data['group'].isin(groups)]
          
          
    returns = \
                  factor_returns(portfolio_data, long_short, group_neutral, equal_weight)
          
          
    return cumulative_returns(returns[period], period)
          
          

          
def factor_positions(factor_data,
                               period,
                               long_short=True,
                               group_neutral=False,
                               equal_weight=False,
                               quantiles=None,
                               groups=None):
              """

Two parameters are passed through but cumulative_returns function only takes one parameter (returns).

Tried to provide feedback via github but was disabled for me.

Cheers

Brian · May 4, 2022, 12:42pm

Quantopian simplified cumulative_returns at some point to only support daily returns, but failed to update this function accordingly.

This is now fixed in master, will be included in the next release.