ENH: a couple notes on IntervalIndex, TimeInterval

pandas/core/datetools.py

@@ -70,6 +70,44 @@ def parser(x):
     data = p_ufunc(arr)
     return np.array(data, dtype='M8[us]')
 
+#-------
+# Interval sketching
+
+_dt_component_ordinal = {
+    "year" : 0,
+    "month" : 2,
+    "day" : 3,
+    "hour" : 4,
+    "minute" : 5,
+    "second" : 6,
+    "microsecond" : 8
+}
+
+_ti_ordinal = {
+    'Y' : 0,
+    'Q' : 1,
+    'M' : 2,
+    'D' : 3,
+    'H' : 4,
+    'm' : 5,
+    's' : 6,
+    'ms' : 7,
+    'us' : 8
+}
+
+class TimeInterval:
+    def __init__(self, ival, freq='D'):
+        ret, parsed, reso = parse_time_string(ival)
+
+        if _dt_component_ordinal[reso] > _ti_ordinal[freq]:
+            pass
+
+def to_interval(val, freq):
+    if isinstance(val, datetime):
+        return val
+    if isinstance(val, int):
+        return TimeInterval(val, freq)
+
 #-------------------------------------------------------------------------------
 # Miscellaneous date functions
 

pandas/core/index.py

@@ -16,7 +16,7 @@
 
 import pandas.core.datetools as datetools
 from pandas.core.datetools import (_dt_box, _dt_unbox, _dt_box_array,
-                                   _dt_unbox_array, to_timestamp)
+                                   _dt_unbox_array, to_timestamp, Interval)
 
 __all__ = ['Index']
 
@@ -1076,7 +1076,6 @@ def _maybe_box_dtindex(idx):
         return Index(_dt_box_array(idx.asi8), dtype='object')
     return idx
 
-
 class DatetimeIndex(Int64Index):
     """
     Immutable ndarray of datetime64 data, represented internally as int64, and
@@ -1154,7 +1153,7 @@ def __new__(cls, data=None,
         offset = freq
 
         if data is None and offset is None:
-            raise ValueError("Must provide offset argument if no data is "
+            raise ValueError("Must provide freq argument if no data is "
                              "supplied")
 
         if data is None:
@@ -1805,6 +1804,79 @@ def tz_validate(self):
 
         return True
 
+# --- Interval index sketch
+
+class IntervalIndex(Int64Index):
+    """
+    Immutable ndarray where values represent interval offsets from the
+    Gregorian proleptic date 1/1/1.
+
+    Intervals are boxed to Interval objects which carry metadata such as
+    frequency information and start and end datetimes.
+
+    Parameters
+    ----------
+    data  : array-like (1-dimensional), optional
+        Optional interval-like data to construct index with
+    dtype : NumPy dtype (default: i8)
+    copy  : bool
+        Make a copy of input ndarray
+    freq : string or interval object, optional
+        One of pandas interval strings or corresponding objects
+    start : starting value, interval-like, optional
+        If data is None, used as the start point in generating regular
+        interval data.
+    periods  : int, optional, > 0
+        Number of intervals to generate, if generating data. Takes precedence
+        over end argument
+    end   : end value, interval-like, optional
+        If periods is none, generated index will extend to first conforming
+        interval on or just past end argument
+    """
+
+    def __new__(cls, data=None,
+                freq=None, start=None, end=None, periods=None,
+                copy=False, name=None):
+
+        if data is None and freq is None:
+            raise ValueError("Must provide freq argument if no data is "
+                             "supplied")
+
+        if data is None:
+            start = datetools.to_interval(start, freq)
+            end = datetools.to_interval(end, freq)
+
+            if (start is not None and not isinstance(start, Interval)):
+                raise ValueError('Failed to convert %s to interval' % start)
+
+            if (end is not None and not isinstance(end, Interval)):
+                raise ValueError('Failed to convert %s to interval' % end)
+
+            data = np.arange(start.value, end.value+1, dtype=np.int64)
+
+            index = data.view(cls)
+            index.name = name
+            index.freq = freq
+
+            return index
+
+        if not isinstance(data, np.ndarray):
+            if np.isscalar(data):
+                raise ValueError('IntervalIndex() must be called with a '
+                                 'collection of some kind, %s was passed'
+                                 % repr(data))
+
+            if isinstance(data, int):
+                data = [data]
+
+            # other iterable of some kind
+            if not isinstance(data, (list, tuple)):
+                data = list(data)
+
+        subarr = data.view(cls)
+        subarr.name = name
+
+        return subarr
 # --------------------------- end of datetime-specific code ---------------
 
 class Factor(np.ndarray):

pandas/tests/test_datetime64.py

@@ -21,6 +21,9 @@
 from pandas.core.groupby import Tinterval
 from pandas.core.datetools import Minute, BDay
 
+from pandas.core.index import IntervalIndex
+from pandas.core.datetools import TimeInterval
+
 try:
     import pytz
 except ImportError:
@@ -558,6 +561,12 @@ def test_datetimeindex_constructor(self):
         for other in [idx2, idx3, idx4, idx5, idx6]:
             self.assert_( (idx1.values == other.values).all() )
 
+    def test_intervalindex_constructor(self):
+        ii = IntervalIndex(freq='M', start='1/1/2005', end='12/1/2005')
+        self.assertEquals(len(ii), 12)
+        self.assertEquals(ii[0], TimeInterval('1/2005', freq='M'))
+        self.assertEquals(ii[-1], TimeInterval('12/2005', freq='M'))
+
 if __name__ == '__main__':
     import nose
     nose.runmodule(argv=[__file__,'-vvs','-x','--pdb', '--pdb-failure'],