j220

I am getting a similar error. Empyrical and all other dependencies are up to date. Return and benchmark series have definitely the same length. Traces back to sklearn. Any solutions...

Thanks for the help @twiecki. I updated but the error is still the same.

Now the error changes: ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

``` ValueError Traceback (most recent call last) in () 2 livestart='2017-01-01' 3 #benchmark_rets=test['returns'], ----> 4 pf.create_returns_tear_sheet(test['dnn_returns'], benchmark_rets=test['returns'], live_start_date=livestart) ~/anaconda/lib/python3.6/site-packages/pyfolio/plotting.py in call_w_context(*args, **kwargs) 50 if set_context: 51 with plotting_context(), axes_style(): --->...

Double checked, no nans or inf in any series. Maybe an issue with the frequency of my data in business days? `data=(data.resample('B').last())`