The main class of the la package is a labeled array, larry. A larry consists of data and labels. The data is stored as a NumPy array and the labels as a list of lists (one list per dimension).
Here's larry in schematic form:
date1 date2 date3 'AAPL' 209.19 207.87 210.11 y = 'IBM' 129.03 130.39 130.55 'DELL' 14.82 15.11 14.94
The larry above is stored internally as a Numpy array and a list of lists:
y.label = [['AAPL', 'IBM', 'DELL'], [date1, date2, date3]] y.x = np.array([[209.19, 207.87, 210.11], [129.03, 130.39, 130.55], [ 14.82, 15.11, 14.94]])
A larry can have any number of dimensions except zero. Here, for example, is one way to create a one-dimensional larry:
>>> import la >>> y = la.larry([1, 2, 3])
In the statement above the list is converted to a Numpy array and the labels
range(n), where n in this case is 3.
larry has built-in methods such as ranking, merge, shuffle, move_sum, zscore, demean, lag as well as typical Numpy methods like sum, max, std, sign, clip. NaNs are treated as missing data.
Alignment by label is automatic when you add (or subtract, multiply, divide) two larrys. You can also specify the join method (inner, outer, left, right) for binary operations on two larrys with unaligned labels.
You can archive larrys in HDF5 format using save and load or using a dictionary-like interface:
>>> io = la.IO('/tmp/dataset.hdf5') >>> io['y'] = y # <--- save >>> z = io['y'] # <--- load >>> del io['y'] # <--- delete from archive
For the most part larry acts like a Numpy array. And, whenever you want, you have direct access to the Numpy array that holds your data. For example if you have a function, myfunc, that works on Numpy arrays and doesn't change the shape or ordering of the array, then you can use it on a larry, y, like this:
y.x = myfunc(y.x)
larry adds the convenience of labels, provides many built-in methods, and let's you use your existing array functions.
la package is distributed under a Simplified BSD license. Parts of
NumPy, Scipy, and numpydoc, which all have BSD licenses, are included in
la. Parts of matplotlib are also included. See the LICENSE file, which
is distributed with the
la package, for details.
- Python 2.7 or 3.2
- NumPy 1.6.2 or 1.7.0
- Bottleneck 0.7.0
- nose (for unit tests)
- h5py, HDF 1.8 (archive larrys in HDF5)
- gcc or MinGW (compile for speed boost)
- SciPy (lar.ranking(norm='gaussian'))
GNU/Linux, Mac OS X et al.
$ python setup.py build $ sudo python setup.py install
Or, if you wish to specify where
la is installed, for example inside
$ python setup.py build $ sudo python setup.py install --prefix=/usr/local
Alternatively, you can use the makefile to install
$ make all
In order to (optionally) compile the C code in the
la package you need a
Windows version of the gcc compiler. MinGW (Minimalist GNU for Windows)
contains gcc and has been used to successfully compile
la on Windows.
Install MinGW and add it to your system path. Then install
la with the
python setup.py build --compiler=mingw32 python setup.py install
After you have installed
la, run the suite of unit tests:
>>> import la >>> la.test() <snip> Ran 2995 tests in 9.225s OK <nose.result.TextTestResult run=2995 errors=0 failures=0>
la package contains C extensions that speed up common alignment
operations such as adding two unaligned larrys. If the C extensions don't
compile when you build
la then there's an automatic fallback to python
versions of the functions. To see whether you are using the C functions or the
>>> la.info() la version 0.7.0 la file /usr/local/lib/python2.6/dist-packages/la/__init__.pyc NumPy 1.6.2 Bottleneck 0.6.0 HDF5 archiving Available (h5py 2.0.0) listmap Faster C version listmap_fill Faster C version
la can run in a pure python mode, you can use
la by just saving
it and making sure that python can find it.