Algorithm Implementation Tips
Implementing algorithms is fun, but there are a few things that are good to keep in mind.
In general, development follows the following:
In that order. Further, we always want LensKit to be usable in an easy fashion. Code implementing algorithms, however, may be quite complex in order to achieve good performance.
We use Numba to optimize critical code paths and provide parallelism in a number of cases, such as ALS training. See the ALS source code for examples.
We also use the CSR package for sparse matrices that are usable from Numba-accelerated code, and to provide unified access to important sparse matrix operations that use MKL acceleration when available. Previous versions of LensKit included the MKL code directly, but we have moved that logic over into CSR.
If you are working on an algorithm implementation that needs access to additional MKL operations, please add the relevant operations to CSR to keep LensKit pure Python + Numba. We do not have plans to re-add the MKL wrapper logic to the LensKit core.
Pickling and Sharing
LensKit uses binpickle quite a bit to save and reload models and to share model data between concurrent processes. This generally just works, and you don’t need to implement any particular save/load logic in order to have your algorithm be savable and sharable.
There are a few exceptions, though.
If your algorithm updates state after fitting, this should not be pickled. An example of
this would be caching predictions or recommendations to save time in subsequent calls. Only the
model parameters and estimated parameters should be pickled. If you have caches or other
ephemeral structures, override
__setstate__ to exclude them from the
saved data and to initialize caches to empty values on unpickling.
If your model excludes secondary data structures from pickling, such as a reverse index of
user-item interactions, then you should only exclude them when pickling for serialization. When
pickling for model sharing (see
lenskit.sharing.in_share_context()), you should include
the derived structures so they can also be shared.
If your algorithm uses subsidiary models as a part of the training process, but does not need them
for prediction or recommendation, then consider overriding
__getstate__ to remove the underlying
model or replace it with a cloned copy (with
lenskit.util.clone()) to reduce serialized
disk space (and deserialized memory use).
Random Number Generation
seedbank for managing RNG seeds and constructing random number generation.
In general, algorithms using randomization should have an
rng_spec parameter that takes a seed
or RNG, and pass this to
seedbank.numpy_rng() to get a random number generator. Algorithms
that use randomness at predict or recommendation time, not just training time, should support the
'user' for the
rng parameter, and if it is passed, derive a new seed for each user
seedbank.derive_seed() to allow reproducibility in the face of parallelism for common
lenskit.util.derivable_rng() automates this logic.