@@ -10,6 +10,43 @@ pandas 0.4.2
10
10
11
11
**Release date: ** not yet released
12
12
13
+ This is an incremental bug fix and performance enhancement release
14
+
15
+ **New features / modules **
16
+
17
+ - Added `Int64Index ` type with specialized join, union, intersection
18
+ - Refactored `Index ` classes to have a `join ` method and associated data
19
+ alignment routines throughout the codebase to be able to leverage optimized
20
+ joining / merging routines.
21
+ - Wrote faster Cython data alignment / merging routines resulting in
22
+ substantial speed increases
23
+ - Added `is_monotonic ` property to `Index ` classes with associated Cython
24
+ code to evaluate the monotonicity of the `Index ` values
25
+
26
+ **Improvements to existing features **
27
+
28
+ - Improved performance of `isnull ` and `notnull `, a regression from v0.3.0
29
+ (GH #187)
30
+ - Wrote templating / code generation script to auto-generate Cython code for
31
+ various functions which need to be available for the 4 major data types
32
+ used in pandas (float64, bool, object, int64)
33
+ - Refactored code related to `DataFrame.join ` so that intermediate aligned
34
+ copies of the data in each `DataFrame ` argument do not need to be
35
+ created. Substantial performance increases result (GH #176)
36
+ - Substantially improved performance of generic `Index.intersection ` and
37
+ `Index.union `
38
+ - Implemented `BlockManager.take ` resulting in significantly faster `take `
39
+ performance on mixed-type `DataFrame ` objects (GH #104)
40
+ - Improved performance of `Series.sort_index `
41
+
42
+ **Bug fixes **
43
+
44
+ - Fixed minor unhandled exception in Cython code implementing fast groupby
45
+ aggregation operations
46
+ - Fixed bug in unstacking code manifesting with more than 3 hierarchical
47
+ levels
48
+ - Throw exception when step specified in label-based slice (GH #185)
49
+
13
50
pandas 0.4.1
14
51
============
15
52
0 commit comments