Consider the following example. python import xarray as xr import numpy as np import dask.array as dsa from dask.diagnostics import ProgressBar ```

Groupby bins problem with small bins?

x_raw = np.arange(20) y_raw = np.arange(10) z_raw = np.arange(15)

x = xr.DataArray(dsa.from_array(x_raw, chunks=(-1)), dims=['x'], coords={'x':('x', x_raw)}) y = xr.DataArray(dsa.from_array(y_raw, chunks=(-1)), dims=['y'], coords={'y':('y', y_raw)}) z = xr.DataArray(dsa.from_array(z_raw, chunks=(-1)), dims=['z'], coords={'z':('z', z_raw)})

data = xr.DataArray(dsa.ones([20, 10, 15], chunks=[-1, -1, -1]), dims=['x', 'y', 'z'], coords={ 'x':x, 'y':y, 'z':z }) data <xarray.DataArray 'wrapped-bb05d395159047b749ca855110244cb7' (x: 20, y: 10, z: 15)> dask.array<shape=(20, 10, 15), dtype=float64, chunksize=(20, 10, 15)> Coordinates: * x (x) int64 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 * y (y) int64 0 1 2 3 4 5 6 7 8 9 * z (z) int64 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

This dask array only contains ones. If I now try to apply groupby_bins with a specified array of bins (which are all below 1) it fails with a rather cryptic error. ```

```

This doesnt work

bins = np.array([0, 20, 40, 60 , 80, 100])*1e-6

binned = data.groupby_bins(data, bins).sum() binned

---

StopIteration Traceback (most recent call last) <ipython-input-7-dc9283bee4ea> in <module>() 2 bins = np.array([0, 20, 40, 60 , 80, 100])*1e-6 3 ----> 4 binned = data.groupby_bins(data, bins).sum() 5 binned

~/Work/CODE/PYTHON/xarray/xarray/core/common.py in wrapped_func(self, dim, axis, skipna, keep_attrs, kwargs) 20 keep_attrs=False, kwargs): 21 return self.reduce(func, dim, axis, keep_attrs=keep_attrs, ---> 22 skipna=skipna, allow_lazy=True, **kwargs) 23 else: 24 def wrapped_func(self, dim=None, axis=None, keep_attrs=False,

~/Work/CODE/PYTHON/xarray/xarray/core/groupby.py in reduce(self, func, dim, axis, keep_attrs, shortcut, kwargs) 572 def reduce_array(ar): 573 return ar.reduce(func, dim, axis, keep_attrs=keep_attrs, kwargs) --> 574 return self.apply(reduce_array, shortcut=shortcut) 575 576 ops.inject_reduce_methods(DataArrayGroupBy)

~/Work/CODE/PYTHON/xarray/xarray/core/groupby.py in apply(self, func, shortcut, kwargs) 516 applied = (maybe_wrap_array(arr, func(arr, kwargs)) 517 for arr in grouped) --> 518 return self._combine(applied, shortcut=shortcut) 519 520 def _combine(self, applied, shortcut=False):

~/Work/CODE/PYTHON/xarray/xarray/core/groupby.py in _combine(self, applied, shortcut) 520 def _combine(self, applied, shortcut=False): 521 """Recombine the applied objects like the original.""" --> 522 applied_example, applied = peek_at(applied) 523 coord, dim, positions = self._infer_concat_args(applied_example) 524 if shortcut:

~/Work/CODE/PYTHON/xarray/xarray/core/utils.py in peek_at(iterable) 114 """ 115 gen = iter(iterable) --> 116 peek = next(gen) 117 return peek, itertools.chain([peek], gen) 118

StopIteration: ```

If however the last bin includes the value 1 it runs as expected: ```

If I include a larger value at the end it works

bins = np.array([0, 20, 40, 60 , 80, 100, 1e7])*1e-6

binned = data.groupby_bins(data, bins).sum() binned <xarray.DataArray 'wrapped-bb05d395159047b749ca855110244cb7' (wrapped-bb05d395159047b749ca855110244cb7_bins: 6)> dask.array<shape=(6,), dtype=float64, chunksize=(5,)> Coordinates: * wrapped-bb05d395159047b749ca855110244cb7_bins (wrapped-bb05d395159047b749ca855110244cb7_bins) object (0.0, 2e-05] ... ```

Problem description

Is this expected behaviour? I would prefer it if it returned nan values for the bins that capture no values. It took me a bit to find out why my script using this was failing, and if this is expected behavior could a more helpful error message be considered?

Expected Output

Output of xr.show_versions()

# Paste the output here xr.show_versions() here INSTALLED VERSIONS ------------------ commit: None python: 3.6.2.final.0 python-bits: 64 OS: Darwin OS-release: 16.7.0 machine: x86_64 processor: i386 byteorder: little LC_ALL: None LANG: en_US.UTF-8 LOCALE: en_US.UTF-8 xarray: 0.10.0rc1-9-gdbf7b01 pandas: 0.20.3 numpy: 1.13.1 scipy: 0.19.1 netCDF4: 1.2.9 h5netcdf: 0.4.1 Nio: None bottleneck: 1.2.1 cyordereddict: None dask: 0.15.4 matplotlib: 2.0.2 cartopy: 0.15.1 seaborn: 0.8.1 setuptools: 36.3.0 pip: 9.0.1 conda: None pytest: 3.2.2 IPython: 6.1.0 sphinx: 1.6.5