Conditional split/groupby

[wesm]

per Andreas Hilboll question on mailing list

Hi,

I have a DataFrame ``DF`` where in one column there's the time (in
hours, as float). I'm looking for an easy way of iterating over
subsets of ``DF`` which are defined by looking for time gaps of 45
minutes or more and using these gaps as seperator between the
different iterators.

As a first shot, I would do something like this::

  diff = np.diff(DF.time)
  idx = np.where(diff>.75)[0]
  DFs = []
  for i in idx:
     DFs.append(DF.ix[from:to]

But maybe there's a simpler solution?

[ghost]

dupe #4059