La

======

*Release date: Not yet released, in development*

**Faster**

- Indexing into larrys with scalars, slices, and 1d numpy arrays
- Binary functions (+, -, / , &, ...) for unaligned larrys
- la.align(), la.add(), la.subtract(), etc for unaligned larrys
- la.union(), la.intersection(), la.align_axis()
- la.farray.correlation(arr1, arr2, axis) for numpy arrays
- larry.merge() when update=True, larry.morph(), larry.morph_like()
- larry.take()
- larry.sortaxis()
- larry.nan_replace()

**Enhancements**

- la now supports Python 3 (thanks to Benjamin Root for initial porting)
- Can now index into a larry with a 1d bool larry
- Added setitem ability to label indexing (lix)
- la.IO no longer keeps the archive connection open between calls
- Can now archive larrys that have tuple labels
- C files generated with Cython 0.21.1 instead of 0.15.1
- misc.fromlists() now raises on overlapping input data

------

*Release date: 2012-03-19*

**New functions**

- la.isaligned() returns True if two larrys are aligned along specified axis
- la.sortby() sorts a larry by a row or column specified by its label
- la.align_axis() aligns multiple larrys along (possibly) different axes
- la.zeros(), la.ones(), la.empty()
- la.lrange() similar to np.arange() but allows multi-dimensional output

**Enhancements**

- larry.lag() now accepts negative lags
- datime.time and datetime.datetime labels can now be (HDF5) archived
- la.align() can now skip the axes you do not wish to align
- Upgrade numpydoc from 0.3.1 to 0.4 to support Sphinx 1.0.1
- la.farray.ranking() and larry ranking method support `axis=None`
- Generate C code with Cython 0.15.1 instead of Cython 0.11
- Add makefile

**Faster**

- larry methods: merge, nan_replace, push, cumsum, cumprod, astype, __rdiv__
- larry function: cov
- Numpy array functions: geometric_mean, correlation, covMissing

------

*Release date: 2011-04-28*

The fifth release of la is faster and adds fast, moving window methods to
larry. The cost of the speed and new functionality is that la now requires the
Bottleneck package (http://pypi.python.org/pypi/Bottleneck).

**Faster**

- sum, mean, std, var, min, max, median, ranking
- correlation, group_median, demean, demedian, zscore

**Moving window**

- fast (Bottleneck): move_sum, move_mean, move_std, move_min, move_max
- slow (Python): move_ranking, move_median, move_func

**New methods and functions**

- la.data.yahoo.quotes(): Given a ticker sequence, return historical Yahoo!
quotes (open, close, high, low, volume) as a 3d larry
- Numpy array functions that ignore NaNs: demean, demedian, and zscore
- la.unique(): Find the unique elements of a larry
- larry.tofile(): Save 1d or 2d larry to text file

**Enhancements**

- Add optional `ddof` input parameter to larry.std() and larry.var()
- Cut import time by 15%

------

*Release date: 2010-07-06*

The focus of this release was binary operations between unaligned larrys with
user control of the join method (five available) and the fill method. A
general binary function, la.binaryop(), was added as were the convenience
functions add, subtract, multiply, divide. Supporting functions such as
la.align(), which aligns two larrys, were also added.

**New larry methods**

- ismissing: A bool larry with element-wise marking of missing values
- take: A copy of the specified elements of a larry along an axis

**New functions**

- rand: Random samples from a uniform distribution
- randn: Random samples from a Gaussian distribution
- missing_marker: Return missing value marker for the given larry
- ismissing: A bool Numpy array with element-wise marking of missing values
- correlation: Correlation of two Numpy arrays along the specified axis
- split: Split into train and test data along given axis
- listmap_fill: Index map a list onto another and index of unmappable elements
- listmap_fill: Cython version of listmap_fill
- align: Align two larrys using one of five join methods
- info: la package information such as version number and HDF5 availability
- binaryop: Binary operation on two larrys with given function and join method
- add: Sum of two larrys using given join and fill methods
- subtract: Difference of two larrys using given join and fill methods
- multiply: Multiply two larrys element-wise using given join and fill methods
- divide: Divide two larrys element-wise using given join and fill methods

**Enhancements**

- listmap now has option to ignore unmappable elements instead of KeyError
- listmap.pyx now has option to ignore unmappable elements instead of KeyError
- larry.morph() is much faster as are methods, such as merge, that use it

------

*Release date: 2010-06-04*

**New larry methods**

- astype: Copy of larry cast to specified type
- geometric_mean: new method based on existing array function

**New functions**

- la.util.resample.cross_validation: k-fold cross validation index iterator
- la.util.resample.bootstrap: bootstrap index iterator
- la.util.misc.listmap: O(n) version of map(list1.index, list2)
- la/src/clistmap.pyx: Cython version of listmap with python fallback

**Enhancements**

- Major performance boost in most larry methods!
- You can now use an optional dtype when creating larrys
- You can now optionally skip the integrity test when creating a new larry
- Add ability to compare (==, >, !=, etc) larrys with lists and tuples
- Documentation and unit tests

------

*Release date: 2010-04-27*

**New larry methods**

- lix : Index into a larry using labels or index numbers or both
- swapaxes : Swap the two specified axes
- sortaxis : Sort data (and label) according to label along specified axis
- flipaxis : Reverse the order of the elements along the specified axis
- tocsv : Save larry to a csv file
- fromcsv : Load a larry from a csv file
- insertaxis : Insert a new axis at the specified position
- invert : Element by element inverting of True to False and False to True

**Enhancements**

- All larry methods can now take nd input arrays (some previously 2d only)
- Added ability to save larrys with datetime.date labels to HDF5
- New function (panel) to convert larry of shape (n, m, k) to shape (m*k, n)
- Expanded documentation
- Over 280 new unit tests; testing easier with new assert_larry_equal function

**Bug fixes**

- 517912: larry([]) == larry([]) raised IndexError
- 518096: larry.fromdict failed due to missing import
- 518106: la.larry.fromdict({}) failed
- 518114: fromlist([]) and fromtuples([]) failed
- 518135: keep_label crashed when there was nothing to keep
- 518210: sum, std, var returned NaN for empty larrys; now return 0.0
- 518215: unflatten crashed on an empty larry
- 518442: sum, std, var returned NaN for shapes that contain zero: (2, 0, 3)
- 568175: larry.std(axis=-1) and var crashed on negative axis input
- 569622: Negative axis input gave wrong output for several larry methods