The AOCL-DA Python package comes with several example scripts. To locate them, the following commands can be run in your Python interpreter.
from aoclda.examples import info
info.examples_path()
info.examples_list()
Alternatively, from your command prompt the command
python -m aocl.examples.info will print the same information.
For existing scikit-learn users, you can “patch” your python code to use AOCL-DA where it is available by inserting the following lines prior to your scikit-learn import lines:
from aoclda.sklearn import skpatch, undo_skpatch
skpatch()
You can switch back to standard scikit-learn using
undo_skpatch()
Note that after calling undo_skpatch, you must reimport scikit-learn.
The skpatch and undo_skpatch functions can also be called with string or list arguments, specifying which scikit-learn package should be patched, for example:
skpatch("PCA")
skpatch(["LinearRegression", "Ridge"])
Alternatively, you can use the aoclda.sklearn module from the command line, without
making any changes to your own code:
python -m aoclda.sklearn your_python_script.py
python -m aoclda.sklearn -m your_python_module
For scikit-learn classes that use n_jobs to specify the number of threads, instead
use the OMP_NUM_THREADS environment variable to control the number of threads used by AOCL-DA.
Note that only a subset of the AOCL-DA functionality is available in this manner.
For a full list of the supported classes and methods, please refer to the API guide at https://docs.amd.com/r/en-US/68552-AOCL-api-guide/AOCL-Data-Analytics.
If, after patching, you attempt to call scikit-learn class member functions which have not been implemented in AOCL-DA,
then a RuntimeError will be thrown.
Note that Python support on Windows is currently experimental with certain functionality
(the nonlinear least squares and LBFGSB solvers) not available.