Volatility weighting NB seems to fail

[twiecki]

--------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-6-3ab88e0b88b5> in <module>()
      2                                                                        weight_function=np.std,
      3                                                                        weight_function_window=126,
----> 4                                                                        inverse_weight=True
      5                                                                       )
      6 to_plot = ['SPY', 'GLD', 'FXE'] + ["port_ret"]

/home/wiecki/working/projects/quant/pyfolio/pyfolio/timeseries.py in portfolio_returns_metric_weighted(holdings_returns, exclude_non_overlapping, weight_function, weight_function_window, inverse_weight, portfolio_rebalance_rule, weight_func_transform)
   1251             holdings_df_weights = holdings_df[transform_columns] / \
   1252                 holdings_df[transform_columns].sum(axis=1)
-> 1253         holdings_df_weights.columns = holdings_cols
   1254         holdings_df = holdings_df.join(holdings_df_weights, rsuffix='_w')
   1255         holdings_df_weighted_rets = np.multiply(

/home/wiecki/miniconda3/lib/python3.4/site-packages/pandas/core/generic.py in __setattr__(self, name, value)
   2255         try:
   2256             object.__getattribute__(self, name)
-> 2257             return object.__setattr__(self, name, value)
   2258         except AttributeError:
   2259             pass

pandas/src/properties.pyx in pandas.lib.AxisProperty.__set__ (pandas/lib.c:44611)()

/home/wiecki/miniconda3/lib/python3.4/site-packages/pandas/core/generic.py in _set_axis(self, axis, labels)
    422 
    423     def _set_axis(self, axis, labels):
--> 424         self._data.set_axis(axis, labels)
    425         self._clear_item_cache()
    426 

/home/wiecki/miniconda3/lib/python3.4/site-packages/pandas/core/internals.py in set_axis(self, axis, new_labels)
   2458         if new_len != old_len:
   2459             raise ValueError('Length mismatch: Expected axis has %d elements, '
-> 2460                              'new values have %d elements' % (old_len, new_len))
   2461 
   2462         self.axes[axis] = new_labels

ValueError: Length mismatch: Expected axis has 2342 elements, new values have 3 elements

[justinlent]

This is working for me on my local branch. are you trying on Master? any package updates that you've done (like the new pandas?) that may be causing it? i'm still on this pandas:

[twiecki]

I'm on 0.17.0 which might be the issue. Can you try upgrading and see if you can reproduce that way?

[justinlent]

cool, yea., I'll clone my current virtual env, and then just update to pandas 0.17.0 to try and isolate it to pandas

[justinlent]

(but i probably won't get to trying this untill later today in the afternoon as I have some meetings in the morning)

[justinlent]

confirmed. It fails on me when upgrading to pandas 0.17.0. ohhh pandas and how you consistently break backcompat... ;)

i think they messed around with how indexes should be set in this new release. let me see if its a quick fix

[twiecki]

Related: pandas-dev/pandas#11336

[justinlent]

@twiecki PR for your review
#173

[twiecki]

http://pandas.pydata.org/pandas-docs/stable/whatsnew.html#removal-of-prior-version-deprecations-changes