Dashboard errors at experiment start

[dconathan]

When first accessing the experiment dashboard (when an experiment is just launched), I get a ton of errors like this, presumably because there is no data yet.

[31mminionworker_1      | [0m[2016-06-23 22:28:54,736: ERROR/MainProcess] Task next.broker.celery_app.tasks.apply_dashboard[5b6586f1-39f5-42b1-b8f7-4bd604dae723] raised unexpected: IndexError('list index out of range',)
[31mminionworker_1      | [0mTraceback (most recent call last):
[31mminionworker_1      | [0m  File "/usr/local/lib/python2.7/dist-packages/celery/app/trace.py", line 240, in trace_task
[31mminionworker_1      | [0m    R = retval = fun(*args, **kwargs)
[31mminionworker_1      | [0m  File "/usr/local/lib/python2.7/dist-packages/celery/app/trace.py", line 438, in __protected_call__
[31mminionworker_1      | [0m    return self.run(*args, **kwargs)
[31mminionworker_1      | [0m  File "/next_backend/next/broker/celery_app/tasks.py", line 80, in apply_dashboard
[31mminionworker_1      | [0m    **args_dict['args']['params'])
[31mminionworker_1      | [0m  File "/next_backend/next/utils/utils.py", line 160, in timed
[31mminionworker_1      | [0m    result = f(*args, **kw)
[31mminionworker_1      | [0m  File "/next_backend/next/apps/AppDashboard.py", line 38, in api_activity_histogram
[31mminionworker_1      | [0m    ax.hist(numerical_timestamps,min(int(1+4*numpy.sqrt(len(numerical_timestamps))),300),alpha=0.5,color='black')
[31mminionworker_1      | [0m  File "/usr/lib/pymodules/python2.7/matplotlib/axes.py", line 8249, in hist
[31mminionworker_1      | [0m    if isinstance(x, np.ndarray) or not iterable(x[0]):
[31mminionworker_1      | [0mIndexError: list index out of range