add info method to dataset (#1176)

add info method to dataset

.travis.yml

@@ -67,7 +67,7 @@ install:
   - python setup.py install
 
 script:
-  - py.test xarray --cov=xarray --cov-report term-missing
+  - py.test xarray --cov=xarray --cov-report term-missing --verbose
 
 after_success:
   - coveralls

doc/api.rst

@@ -411,6 +411,7 @@ Dataset methods
    Dataset.load
    Dataset.chunk
    Dataset.filter_by_attrs
+   Dataset.info
 
 DataArray methods
 -----------------

doc/whats-new.rst

@@ -158,6 +158,10 @@ Enhancements
   plots (:issue:`897`). See :ref:`plotting.figsize` for more details.
   By `Stephan Hoyer <https://github.com/shoyer>`_ and
   `Fabien Maussion <https://github.com/fmaussion>`_.
+- New :py:meth:`~Dataset.info` method to summarize ``Dataset`` variables
+  and attributes. The method prints to a buffer (e.g. ``stdout``) with output
+  similar to what the command line utility ``ncdump -h`` produces (:issue:`1150`).
+  By `Joe Hamman <https://github.com/jhamman>`_.
 
 Bug fixes
 ~~~~~~~~~

setup.cfg

@@ -1,5 +1,5 @@
 [wheel]
 universal = 1
 
-[pytest]
+[tool:pytest]
 python_files=test_*.py

xarray/core/dataset.py

@@ -5,6 +5,8 @@
 from collections import Mapping
 from numbers import Number
 
+import sys
+
 import numpy as np
 import pandas as pd
 
@@ -25,10 +27,10 @@
 from .variable import (Variable, as_variable, IndexVariable, broadcast_variables)
 from .pycompat import (iteritems, basestring, OrderedDict,
                        dask_array_type, range)
+from .formatting import ensure_valid_repr
 from .combine import concat
 from .options import OPTIONS
 
-
 # list of attributes of pd.DatetimeIndex that are ndarrays of time info
 _DATETIMEINDEX_COMPONENTS = ['year', 'month', 'day', 'hour', 'minute',
                              'second', 'microsecond', 'nanosecond', 'date',
@@ -802,6 +804,43 @@ def to_netcdf(self, path=None, mode='w', format=None, group=None,
     def __unicode__(self):
         return formatting.dataset_repr(self)
 
+    def info(self, buf=None):
+        """
+        Concise summary of a Dataset variables and attributes.
+
+        Parameters
+        ----------
+        buf : writable buffer, defaults to sys.stdout
+
+        See Also
+        --------
+        pandas.DataFrame.assign
+        netCDF's ncdump
+        """
+
+        if buf is None:  # pragma: no cover
+            buf = sys.stdout
+
+        lines = []
+        lines.append(u'xarray.Dataset {')
+        lines.append(u'dimensions:')
+        for name, size in self.dims.items():
+            lines.append(u'\t{name} = {size} ;'.format(name=name, size=size))
+        lines.append(u'\nvariables:')
+        for name, da in self.variables.items():
+            dims = u', '.join(da.dims)
+            lines.append(u'\t{type} {name}({dims}) ;'.format(
+                type=da.dtype, name=name, dims=dims))
+            for k, v in da.attrs.items():
+                lines.append(u'\t\t{name}:{k} = {v} ;'.format(name=name, k=k,
+                                                              v=v))
+        lines.append(u'\n// global attributes:')
+        for k, v in self.attrs.items():
+            lines.append(u'\t:{k} = {v} ;'.format(k=k, v=v))
+        lines.append(u'}')
+
+        buf.write(u'\n'.join(lines))
+
     @property
     def chunks(self):
         """Block dimensions for this dataset's data or None if it's not a dask

xarray/core/formatting.py

@@ -16,7 +16,7 @@
 
 from .options import OPTIONS
 from .pycompat import (
-    PY2, unicode_type, bytes_type, dask_array_type, OrderedDict)
+    PY2, unicode_type, bytes_type, dask_array_type, OrderedDict, basestring)
 
 
 def pretty_print(x, numchars):
@@ -87,6 +87,7 @@ def first_n_items(x, n_desired):
         x = x[indexer]
     return np.asarray(x).flat[:n_desired]
 
+
 def last_item(x):
     """Returns the last item of an array"""
     if x.size == 0:
@@ -96,6 +97,7 @@ def last_item(x):
     indexer = (slice(-1, None), ) * x.ndim
     return np.array(x[indexer], ndmin=1)
 
+
 def format_timestamp(t):
     """Cast given object to a Timestamp and return a nicely formatted string"""
     # Timestamp is only valid for 1678 to 2262

xarray/test/test_dataset.py

@@ -12,6 +12,7 @@
     import dask.array as da
 except ImportError:
     pass
+from io import StringIO
 
 import numpy as np
 import pandas as pd
@@ -190,6 +191,42 @@ def test_unicode_data(self):
         actual = unicode_type(data)
         self.assertEqual(expected, actual)
 
+    def test_info(self):
+        ds = create_test_data(seed=123)
+        ds = ds.drop('dim3')  # string type prints differently in PY2 vs PY3
+        ds.attrs['unicode_attr'] = u'ba®'
+        ds.attrs['string_attr'] = 'bar'
+
+        buf = StringIO()
+        ds.info(buf=buf)
+
+        expected = dedent(u'''\
+        xarray.Dataset {
+        dimensions:
+        	dim1 = 8 ;
+        	dim2 = 9 ;
+        	dim3 = 10 ;
+        	time = 20 ;
+
+        variables:
+        	datetime64[ns] time(time) ;
+        	float64 dim2(dim2) ;
+        	float64 var1(dim1, dim2) ;
+        		var1:foo = variable ;
+        	float64 var2(dim1, dim2) ;
+        		var2:foo = variable ;
+        	float64 var3(dim3, dim1) ;
+        		var3:foo = variable ;
+        	int64 numbers(dim3) ;
+
+        // global attributes:
+        	:unicode_attr = ba® ;
+        	:string_attr = bar ;
+        }''')
+        actual = buf.getvalue()
+        self.assertEqual(expected, actual)
+        buf.close()
+
     def test_constructor(self):
         x1 = ('x', 2 * np.arange(100))
         x2 = ('x', np.arange(1000))

xarray/test/test_formatting.py

@@ -87,7 +87,6 @@ def test_format_items(self):
             actual = ' '.join(formatting.format_items(item))
             self.assertEqual(expected, actual)
 
-
     def test_format_array_flat(self):
         actual = formatting.format_array_flat(np.arange(100), 13)
         expected = '0 1 2 3 4 ...'
@@ -126,7 +125,7 @@ def test_format_timestamp_out_of_bounds(self):
         expected = '1300-12-01'
         result = formatting.format_timestamp(date)
         self.assertEqual(result, expected)
-        
+
         date = datetime(2300, 12, 1)
         expected = '2300-12-01'
         result = formatting.format_timestamp(date)