Cannot unpickle `DataFrame` created from period_range

[bjonen]

Running the following snippet on pandas version 0.10.1 leads the traceback at the bottom:

import numpy as np
import pandas as pd

from pandas import *
prng = period_range('1/1/2011', '1/1/2012', freq='M')
ts = Series(np.random.randn(len(prng)), prng)
ts.save('pd_save')
new_ts = pd.load('pd_save')
print new_ts

ValueError                                Traceback (most recent call last)
C:\Python27\lib\site-packages\IPython\utils\py3compat.pyc in execfile(fname, glob, loc)
    169             else:
    170                 filename = fname
--> 171             exec compile(scripttext, filename, 'exec') in glob, loc
    172     else:
    173         def execfile(fname, *where):

C:\development\eclipse-workspace_23\slquant-dev\src\pd_bug.py in <module>()
      8 ts.save('pd_save')
      9 new_ts = pd.load('pd_save')
---> 10 print new_ts

C:\Python27\lib\site-packages\pandas\core\series.pyc in __str__(self)
    968         if py3compat.PY3:
    969             return self.__unicode__()
--> 970         return self.__bytes__()
    971
    972     def __bytes__(self):

C:\Python27\lib\site-packages\pandas\core\series.pyc in __bytes__(self)
    978         """
    979         encoding = com.get_option("display.encoding")
--> 980         return self.__unicode__().encode(encoding, 'replace')
    981
    982     def __unicode__(self):

C:\Python27\lib\site-packages\pandas\core\series.pyc in __unicode__(self)
    994             result = self._get_repr(print_header=True,
    995                                     length=len(self) > 50,
--> 996                                     name=True)
    997         else:
    998             result = com.pprint_thing(self)

C:\Python27\lib\site-packages\pandas\core\series.pyc in _get_repr(self, name, print_header, length, na_rep, float_format)
   1080                                         length=length, na_rep=na_rep,
   1081                                         float_format=float_format)
-> 1082         result = formatter.to_string()
   1083         assert type(result) == unicode
   1084         return result

C:\Python27\lib\site-packages\pandas\core\format.pyc in to_string(self)
    124             return u''
    125
--> 126         fmt_index, have_header = self._get_formatted_index()
    127         fmt_values = self._get_formatted_values()
    128

C:\Python27\lib\site-packages\pandas\core\format.pyc in _get_formatted_index(self)
    110         else:
    111             have_header = index.name is not None
--> 112             fmt_index = index.format(name=True)
    113         return fmt_index, have_header
    114

C:\Python27\lib\site-packages\pandas\tseries\period.pyc in format(self, name, formatter)
   1006             header.append(str(self.name) if self.name is not None else '')
   1007
-> 1008         return header + ['%s' % Period(x, freq=self.freq) for x in self]
   1009
   1010     def __array_finalize__(self, obj):

C:\Python27\lib\site-packages\pandas\tseries\period.pyc in __iter__(self)
    683     def __iter__(self):
    684         for val in self.values:
--> 685             yield Period(ordinal=val, freq=self.freq)
    686
    687     @property

C:\Python27\lib\site-packages\pandas\tseries\period.pyc in __init__(self, value, freq, ordinal, year, month, quarter, day, hour, minute, second)
     79                 raise ValueError("Ordinal must be an integer")
     80             if freq is None:
---> 81                 raise ValueError('Must supply freq for ordinal value')
     82             self.ordinal = ordinal
     83

ValueError: Must supply freq for ordinal value

Fixed in master by #3224. Thanks for reporting this.