v0.5.7

• Fix resource reload, use GET instead of HEAD method
• Fix bugs and structures adjustment
• cloudpickle now support pickle py3 code to py2

v0.5.6

• fix switch bug for DataFrame pandas backend
• fix models list bugs
• secret mode to keep access key safe

v0.5.5

• security API
• fix several DataFrame bugs
• DataFrame SQL backend pypy support

v0.5.4

• add unknown_as_string and as_type parameters when initializing a DataFrame from pandas.DataFrame
• fix table open_reader and open_writer bug when reopen a different partition reader or writer

v0.5.3

• Better tunnel performance(cython required)
• Dataframe bug fixes

v0.5.2

• C integration for Tunnel Download which has better performance
• fix bug of count expression for pandas backend
• fix installation issue

v0.5.0

• User defined aggregation support
• User defined function like map, apply, agg and map_reduce can read resources from file, table or collection
• Window function support for pandas backend
• C integration for Tunnel Upload which has better performance
• Visualization for DataFrame which need the installation of graphviz
• sample API
• strptime for string operation
• bloom_filter and hll_count API
• BuilinFunction for DataFrame to call the builtin function of ODPS SQL which has no parameter or constant parameters
• Display option will sync to pandas option
• Refactoring and bugfixes

v0.4.8

• support aliases for execute_sql
• fix radd bug for DataFrame arithmetic operation
• fix volume api

v0.4.5

• fix bug when creating a table from existed table's schema
• fix bug of df

v0.4.3

• Fix bug of MapReduce API, add ascending parameter.
• Fix bug of sql magic, when ';' exists in sql, error will be raised.