cornac

library2.5.0pypypi✓ verified 85d ago

Cornac is a comparative framework for multimodal recommender systems, providing a rich set of algorithms and evaluation metrics. Version 2.5.0 requires Python >= 3.10. Released under Apache 2.0, with irregular release cadence.

ai-ml

pip install cornac

INSTALL

IMPORT

SIG · CORNAC

cornac

ai-mlpythonv2.5.0

harness data pending

Install & Compatibility

Where this runs

No compatibility data collected yet for this library.

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

Cornac

✓ import cornac

✗ import cornac as cornac

Common mistake: using uppercase C, but module is lowercase

RatioSplit

✓ from cornac.data import Reader from cornac.data import RatioSplit

✗ from cornac import RatioSplit

RatioSplit is in cornac.data, not top-level

Minimal example: load data, split, train PMF, evaluate with RMSE/MAE.

import cornac from cornac.data import Reader, RatioSplit from cornac.models import PMF, BPR, MostPop from cornac.eval_methods import BaseMethod from cornac.metrics import RMSE, MAE, Precision, Recall, NDCG # Load MovieLens 100K reader = Reader() train, test = reader.load('ml-100k', reader=Reader, min_user_freq=5, min_item_freq=2) # Evaluation method ratio_split = RatioSplit(data=train, test_size=0.2, rating_threshold=4.0, seed=42, exclude_unknowns=True) # Models pmf = PMF(k=10, max_iter=50, learning_rate=0.001, lambda_reg=0.001) # Evaluate cornac.Experiment(eval_method=ratio_split, models=[pmf], metrics=[RMSE(), MAE()]).run()

Debug

Known issues

breakingPython 3.10+ only: Cornac 2.5.0 drops support for Python 3.8 and 3.9.

fix

Upgrade Python to 3.10 or later, or install an older version (e.g., pip install cornac==2.4.2)

affects: >=2.5.0

deprecatedThe `cornac.data.Dataset` class is deprecated in favor of `cornac.data.Reader` for loading data.

fix

Use `Reader().load(...)` instead of `Dataset(...)`

affects: >=2.4.0

gotchaImplicit rating threshold: When using `RatioSplit`, the `rating_threshold` parameter is required for binary relevance metrics (e.g., Precision, Recall). Forgetting it will cause errors or incorrect results.

fix

Always specify `rating_threshold` (e.g., 4.0) when using classification metrics.

affects: all

Errors

Common errors & fixes

ModuleNotFoundError: No module named 'cornac'

Cornac not installed in the current Python environment.

fix

Run `pip install cornac`

ImportError: cannot import name 'RatioSplit' from 'cornac'

RatioSplit is not in the top-level module; it's in cornac.data.

fix

Use `from cornac.data import RatioSplit`

ValueError: rating_threshold must be specified for binary classification

Missing `rating_threshold` when using metrics like Precision, Recall.

fix

Add `rating_threshold=4.0` in `RatioSplit` or `BaseMethod`.

Upgrade

Version history

2.5.0latest on PyPI · released Jun 4, 2026

Audit

Dependencies

numpyrequiredCore requirement for numerical operations

scipyrequiredSparse matrix support

tensorflowoptionalDeep learning models

pytorchoptionalAlternative deep learning backend

Agent activity

19 hits · last 30 days

node

Amazon

OpenAI (training)

Resources

No resource links recorded.