Truncate array repr based on line count

[max-sixty]

MCVE Code Sample

I thought we might have had an issue (and maybe solved it?) but couldn't find it anywhere. Forgive me if I'm duplicating.

xr.DataArray(np.random.rand(100,5,1))

<xarray.DataArray (dim_0: 100, dim_1: 5, dim_2: 1)>
array([[[0.71333665],
        [0.93820892],
        [0.48678056],
        [0.07299961],
        [0.63542414]],

*** Deleted 400 lines ***

       [[0.29987457],
        [0.55963998],
        [0.25976744],
        [0.80062955],
        [0.503025  ]],

       [[0.48255097],
        [0.55861315],
        [0.36059861],
        [0.96539665],
        [0.05674621]],

       [[0.81389941],
        [0.55745028],
        [0.20348983],
        [0.63390148],
        [0.94698865]],

       [[0.16792246],
        [0.9252646 ],
        [0.38596734],
        [0.17168077],
        [0.18162088]],

       [[0.04526339],
        [0.70028912],
        [0.72388995],
        [0.97481276],
        [0.66155381]],

       [[0.15058745],
        [0.57646963],
        [0.53382085],
        [0.24696459],
        [0.77601528]],

       [[0.6752243 ],
        [0.84991466],
        [0.87758404],
        [0.70828751],
        [0.04033709]]])
Dimensions without coordinates: dim_0, dim_1, dim_2

Expected Output

With larger arrays, it's much more reasonable:

<xarray.DataArray (dim_0: 500, dim_1: 6, dim_2: 1)>
array([[[0.9680447 ],
        [0.12554914],
        [0.9163406 ],
        [0.63710986],
        [0.97778361],
        [0.6419909 ]],

       [[0.48480678],
        [0.31214637],
        [0.72270997],
        [0.81523543],
        [0.34327902],
        [0.80941523]],

       [[0.92192284],
        [0.47841933],
        [0.00760903],
        [0.83886152],
        [0.88538772],
        [0.6532889 ]],

       ...,

       [[0.39558324],
        [0.42220218],
        [0.56731915],
        [0.27388751],
        [0.51097741],
        [0.62824705]],

       [[0.97379019],
        [0.0311196 ],
        [0.09790975],
        [0.65206508],
        [0.14369363],
        [0.09683937]],

       [[0.71318171],
        [0.88591664],
        [0.30032286],
        [0.97324135],
        [0.10250702],
        [0.03973667]]])
Dimensions without coordinates: dim_0, dim_1, dim_2

Problem Description

Something like 40 lines is probably a reasonable place to truncate?

Output of xr.show_versions()

# Paste the output here xr.show_versions() here INSTALLED VERSIONS ------------------ commit: None python: 3.7.3 | packaged by conda-forge | (default, Jul 1 2019, 21:52:21) [GCC 7.3.0] python-bits: 64 OS: Linux OS-release: [...] machine: x86_64 processor: x86_64 byteorder: little LC_ALL: None LANG: en_US.utf8 LOCALE: en_US.UTF-8 libhdf5: 1.10.5 libnetcdf: 4.7.1

xarray: 0.15.0
pandas: 0.25.3
numpy: 1.17.3
scipy: 1.3.2
netCDF4: 1.5.3
pydap: None
h5netcdf: 0.7.4
h5py: 2.10.0
Nio: None
zarr: None
cftime: 1.0.4.2
nc_time_axis: None
PseudoNetCDF: None
rasterio: None
cfgrib: None
iris: None
bottleneck: 1.2.1
dask: 2.7.0
distributed: 2.7.0
matplotlib: 3.1.2
cartopy: None
seaborn: 0.9.0
numbagg: None
setuptools: 41.6.0.post20191101
pip: 19.3.1
conda: None
pytest: 5.2.2
IPython: 7.9.0
sphinx: 2.2.1

[shoyer]

Hmm, that does look pretty bad. We do have some existing heuristics for shortening up NumPy’s repr but perhaps they could use some tweaking:
https://github.com/pydata/xarray/blob/master/xarray/core/formatting.py#L416