mrec.parallel Package

item_similarity Module

class mrec.parallel.item_similarity.ItemSimilarityRunner

Bases: object

Methods

create_tasks(model, input_format, trainfile, ...)
find_done(outdir)
run(view, model, input_format, trainfile, ...)

create_tasks(model, input_format, trainfile, outdir, num_items, num_engines, max_similar_items, done)

find_done(outdir)

run(view, model, input_format, trainfile, num_engines, simsdir, overwrite, max_sims, simsfile, modelfile)

mrec.parallel.item_similarity.process(task)

Training task to run on an ipython engine.

wrmf Module

class mrec.parallel.wrmf.WRMFRunner

Bases: object

Methods

create_tasks(num_factors, num_engines, ...)
run(view, model, input_format, trainfile, ...)
run_tasks(view, tasks)

create_tasks(num_factors, num_engines, model, input_format, trainfile, workdir, factor_type, get_indices, get_fixed_factor_files, init_fixed_factors)

run(view, model, input_format, trainfile, num_engines, workdir, modelfile)

run_tasks(view, tasks)

mrec.parallel.wrmf.compute_factors(task)

WRMF update method to run on an IPython engine. This reads from file and writes back to file, only filepaths and an empty model need to be passed.

mrec.parallel.wrmf.get_factor_files(workdir, factor_type)

mrec.parallel.wrmf.get_item_factor_files(workdir)

mrec.parallel.wrmf.get_item_indices(data, i)

mrec.parallel.wrmf.get_user_factor_files(workdir)

mrec.parallel.wrmf.get_user_indices(data, u)

mrec.parallel.wrmf.init_item_factors(model, data)

warp Module

class mrec.parallel.warp.WARPMFRunner

Bases: object

Methods

create_tasks(model, input_format, trainfile, ...)
find_done(outdir)
get_modelfile(ix, workdir)
run(view, model, input_format, trainfile, ...)

create_tasks(model, input_format, trainfile, feature_format, featurefile, outdir, num_engines, done)

find_done(outdir)

get_modelfile(ix, workdir)

run(view, model, input_format, trainfile, feature_format, featurefile, num_engines, workdir, overwrite, modelfile)

mrec.parallel.warp.process(task)

Training task to run on an ipython engine.

predict Module

Prediction task to run on an ipython engine.

mrec.parallel.predict.run(task)