check _is_monthly()

[tramsauer]

if there's a gap in the time series (e.g. monthly values for 1980, 1982, 1983, ..) the test returns false.
For calculation of climatology (e.g. get_climatology) this could possibly be neglected (?) as it also breaks the automatic creation of self.time_cycle and hence hinders the calculation.

Possible solutions:

• when checking for increasing time series in _is_monthly()(also daily, yearly):
  • look at combination of month & years in np.diff (out[n] = a[n+1] - a[n]) and only allow positive values.
• add check if time series is continuous