Change naming from ordcomp to cblearn (cblearn#12)

* Find and replace ordcomp by cblearn

* Rename module directory ordcomp to cblearn

* Set main as base branch in CI

.github/workflows/python-package.yml

@@ -5,9 +5,9 @@ name: Tests
 
 on:
   push:
-    branches: [ master ]
+    branches: [ main ]
   pull_request:
-    branches: [ master ]
+    branches: [ main ]
 
 jobs:
   build:
@@ -37,7 +37,7 @@ jobs:
         flake8 . --count --statistics
     - name: Test with pytest and measure coverage
       run: |
-        pytest ordcomp --cov=ordcomp --cov-report=xml
+        pytest cblearn --cov=cblearn --cov-report=xml
     - name: Upload coverage to Codecov
       uses: codecov/codecov-action@v1
       with:
@@ -47,4 +47,4 @@ jobs:
         env_vars: OS,PYTHON
     - name: Build the documentation
       run: |
-        python setup.py build_sphinx
+        python setup.py build_sphinx

LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2020-2021 The ordcomp developers.
+Copyright (c) 2020-2021 The cblearn developers.
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

README.md

@@ -1,9 +1,9 @@
-# ordcomp
-Machine Learning Toolkit for Comparison-based Learning in Python
+# cblearn
+## Comparison-based Machine Learning in Python
 
-[![Unit tests](https://github.com/dekuenstle/ordcomp/workflows/Python%20package/badge.svg)](https://github.com/dekuenstle/ordcomp/actions)
-[![Test Coverage](https://codecov.io/gh/dekuenstle/ordcomp/branch/master/graph/badge.svg?token=P9JRT6OK6O)](https://codecov.io/gh/dekuenstle/ordcomp)
-[![Documentation](https://readthedocs.org/projects/ordcomp/badge/?version=latest)](https://ordcomp.readthedocs.io/en/latest/?badge=latest)
+[![Unit tests](https://github.com/dekuenstle/cblearn/workflows/Python%20package/badge.svg)](https://github.com/dekuenstle/cblearn/actions)
+[![Test Coverage](https://codecov.io/gh/dekuenstle/cblearn/branch/master/graph/badge.svg?token=P9JRT6OK6O)](https://codecov.io/gh/dekuenstle/cblearn)
+[![Documentation](https://readthedocs.org/projects/cblearn/badge/?version=latest)](https://cblearn.readthedocs.io/en/latest/?badge=latest)
 
 Comparison-based Learning are the Machine Learning algorithms to use, when training data
 are ordinal comparisons instead of Euclidean points. 
@@ -16,9 +16,9 @@ It plays hand-in-hand with scikit-learn:
 from sklearn.datasets import load_iris
 from sklearn.model_selection import cross_val_score
 
-from ordcomp.datasets import make_random_triplets
-from ordcomp.embedding.wrapper import SOE
-from ordcomp.metrics import TripletScorer
+from cblearn.datasets import make_random_triplets
+from cblearn.embedding.wrapper import SOE
+from cblearn.metrics import TripletScorer
 
 
 X = load_iris().data
@@ -32,25 +32,25 @@ embedding = estimator.fit_transform(triplets, responses)
 print(f"The embedding has shape {embedding.shape}.")
 ```
 
-Please try the [Examples](https://ordcomp.readthedocs.io/en/latest/generated_examples/index.html).
+Please try the [Examples](https://cblearn.readthedocs.io/en/latest/generated_examples/index.html).
 
 ## Getting Started
 
-OrdComp required Python 3.7 or newer. The easiest way to install is using `pip`:
+cblearn required Python 3.7 or newer. The easiest way to install is using `pip`:
 
 ```
-pip install https://github.com/dekuenstle/ordcomp.git
+pip install https://github.com/dekuenstle/cblearn.git
 ```
-Find more details in the [installation instructions](https://ordcomp.readthedocs.io/en/latest/install.html).
+Find more details in the [installation instructions](https://cblearn.readthedocs.io/en/latest/install.html).
 
 
-In the [User Guide](https://ordcomp.readthedocs.io/en/latest/user_guide/index.html) you find a detailed introduction.
+In the [User Guide](https://cblearn.readthedocs.io/en/latest/user_guide/index.html) you find a detailed introduction.
 
 ## Contribute
 
 We are happy about your contributions.
-Please see our [Contributor Guide](https://ordcomp.readthedocs.io/en/latest/contributor_guide/index.html). 
+Please see our [Contributor Guide](https://cblearn.readthedocs.io/en/latest/contributor_guide/index.html). 
 
 ## License
 
-This library is free to use under the [MIT License](https://github.com/dekuenstle/ordcomp/blob/master/LICENSE) conditions.
+This library is free to use under the [MIT License](https://github.com/dekuenstle/cblearn/blob/master/LICENSE) conditions.

ordcomp/datasets/_triplet_answers.py → cblearn/datasets/_triplet_answers.py

@@ -6,7 +6,7 @@
 from sklearn.metrics import pairwise
 import numpy as np
 
-from ordcomp import utils
+from cblearn import utils
 
 
 class NoiseTarget(enum.Enum):
@@ -48,7 +48,7 @@ def noisy_triplet_answers(triplets: utils.Questions, embedding: np.ndarray, resu
         If return_indices is True, a tuple of indices and answers can be returned
 
 
-    >>> from ordcomp.datasets import noisy_triplet_answers
+    >>> from cblearn.datasets import noisy_triplet_answers
     >>> triplets = [[0, 1, 2], [1, 2, 3]]
     >>> embedding = [[0.1], [0.5], [0.9], [1.]]
     >>> noisy_triplet_answers(triplets, embedding, result_format='list-order')

ordcomp/datasets/_triplet_simulation.py → cblearn/datasets/_triplet_simulation.py

@@ -20,9 +20,9 @@ def make_all_triplets(embedding: np.ndarray, result_format: str, monotonic: bool
             embedding: Object coordinates or distance matrix
             monotonic: Only triplets (j, i, k), such that j < i < k.
             random_state: Seed for noisy answers
-            kwargs: Additional arguments passed to :func:`ordcomp.datasets.noisy_triplet_answers`
+            kwargs: Additional arguments passed to :func:`cblearn.datasets.noisy_triplet_answers`
         Returns:
-            The triplets and answers, based on format. See :func:`ordcomp.utils.check_triplets`.
+            The triplets and answers, based on format. See :func:`cblearn.utils.check_triplets`.
     """
     triplets = make_all_triplet_indices(len(embedding), monotonic)
     return noisy_triplet_answers(triplets, embedding, result_format=result_format, **kwargs)
@@ -49,9 +49,9 @@ def make_random_triplets(embedding: np.ndarray, result_format: str, size: Union[
             make_all: Choose from all triplets instead of iterative sampling,
                       if the difference between all triplets to the requested number is smaller than this value.
             random_state: Seed for triplet sampling and noisy answers
-            kwargs: Additional arguments passed to :func:`ordcomp.datasets.noisy_triplet_answers`
+            kwargs: Additional arguments passed to :func:`cblearn.datasets.noisy_triplet_answers`
         Returns:
-            The triplets and answers, based on format. See :func:`ordcomp.utils.check_triplets`.
+            The triplets and answers, based on format. See :func:`cblearn.utils.check_triplets`.
     """
     triplets = make_random_triplet_indices(len(embedding), size, random_state, repeat, monotonic, make_all)
     return noisy_triplet_answers(triplets, embedding, result_format=result_format, random_state=random_state, **kwargs)

ordcomp/datasets/descr/food_similarity.rst → cblearn/datasets/descr/food_similarity.rst

@@ -19,7 +19,7 @@ Per user selection, multiple triplet constraints were created.
     Dimensionality                      unknown
     ===================   =====================
 
-This dataset can be downloaded using the :func:`ordcomp.datasets.fetch_food_similarity`.
+This dataset can be downloaded using the :func:`cblearn.datasets.fetch_food_similarity`.
 
 .. License statement from the original homepage
 

ordcomp/datasets/descr/musician_similarity.rst → cblearn/datasets/descr/musician_similarity.rst

@@ -24,7 +24,7 @@ Such, for each user judgement multiple triplets were created with the remaining
     The original dataset, published 2002-10-15, contains 224793 triplets.
     We omit in this dataset the triplets with missing values for the last triplet index.
 
-This dataset can be downloaded using the :func:`ordcomp.datasets.fetch_musician_similarity`.
+This dataset can be downloaded using the :func:`cblearn.datasets.fetch_musician_similarity`.
 
 When using these triplets, please give credit to the original authors.
 

ordcomp/datasets/tests/test_food_similarity.py → cblearn/datasets/tests/test_food_similarity.py

@@ -1,10 +1,10 @@
 import numpy as np
 
-from ordcomp.datasets import fetch_food_similarity
+from cblearn.datasets import fetch_food_similarity
 
 
 def test_fetch_food(tmp_path):
-    data_home = tmp_path / 'ordcomp_datasets'
+    data_home = tmp_path / 'cblearn_datasets'
     bunch = fetch_food_similarity(data_home=data_home, shuffle=False)
 
     assert bunch.data.shape == (190376, 3)

ordcomp/datasets/tests/test_musician_similarity.py → cblearn/datasets/tests/test_musician_similarity.py

@@ -1,10 +1,10 @@
 import numpy as np
 
-from ordcomp.datasets import fetch_musician_similarity
+from cblearn.datasets import fetch_musician_similarity
 
 
 def test_fetch_musician_similarity(tmp_path):
-    data_home = tmp_path / 'ordcomp_datasets'
+    data_home = tmp_path / 'cblearn_datasets'
     bunch = fetch_musician_similarity(data_home=data_home, shuffle=False)
 
     assert bunch.data.shape == (213629, 3)

ordcomp/datasets/tests/test_triplet_indices.py → cblearn/datasets/tests/test_triplet_indices.py

@@ -1,7 +1,7 @@
 import numpy as np
 import pytest
 
-from ordcomp import datasets
+from cblearn import datasets
 
 
 def test_make_triplets_raises():

cblearn/embedding/__init__.py

@@ -0,0 +1 @@
+from cblearn.embedding._soe import SOE

ordcomp/embedding/_base.py → cblearn/embedding/_base.py

@@ -4,8 +4,8 @@
 from sklearn.base import TransformerMixin
 from sklearn.utils.validation import check_is_fitted
 
-from ordcomp import datasets
-from ordcomp import utils
+from cblearn import datasets
+from cblearn import utils
 
 
 class TripletEmbeddingMixin(TransformerMixin):

ordcomp/embedding/_soe.py → cblearn/embedding/_soe.py

@@ -6,10 +6,9 @@
 from scipy.optimize import minimize
 from scipy.spatial import distance_matrix
 
-import ordcomp
-from ordcomp import utils
-from ordcomp.embedding._base import TripletEmbeddingMixin
-from ordcomp.utils import assert_torch_is_available, torch_minimize_lbfgs
+from cblearn import utils
+from cblearn.embedding._base import TripletEmbeddingMixin
+from cblearn.utils import assert_torch_is_available, torch_minimize_lbfgs
 
 
 class SOE(BaseEstimator, TripletEmbeddingMixin):
@@ -31,7 +30,7 @@ class SOE(BaseEstimator, TripletEmbeddingMixin):
 
         Examples:
 
-        >>> from ordcomp import datasets
+        >>> from cblearn import datasets
         >>> true_embedding = np.random.rand(15, 2)
         >>> triplets = datasets.make_random_triplets(true_embedding, result_format='list-order', size=1000)
         >>> triplets.shape, np.unique(triplets).shape
@@ -91,7 +90,7 @@ def __init__(self, n_components=2, margin=1, max_iter=1000, verbose=False,
         self.device = device
 
     def fit(self, X: utils.Questions, y: np.ndarray = None, init: np.ndarray = None,
-            n_objects: Optional[int] = None) -> 'ordcomp.embedding.SOE':
+            n_objects: Optional[int] = None) -> 'SOE':
         """Computes the embedding.
 
         Args:

ordcomp/embedding/tests/test_sklearn_estimator_checks.py → cblearn/embedding/tests/test_sklearn_estimator_checks.py

@@ -7,7 +7,7 @@
 This is why we wrap the ordinal embedding estimators below
 to transform featurized input to triplets.
 
-Some checks are skipped, because ordcomp's estimators are not
+Some checks are skipped, because cblearn's estimators are not
 100% compatible to sklearn's estimators.
 While some of the checks could be adapted to our setting,
 some cannot work with triplet input.
@@ -18,9 +18,9 @@
 import numpy as np
 from sklearn.utils.estimator_checks import parametrize_with_checks
 
-from ordcomp.embedding import SOE
-from ordcomp.embedding import wrapper
-from ordcomp.datasets import make_random_triplets
+from cblearn.embedding import SOE
+from cblearn.embedding import wrapper
+from cblearn.datasets import make_random_triplets
 
 
 # Add new estimators here:
@@ -81,7 +81,7 @@ def wrap_triplet_estimator(estimator):
 )
 def test_all_estimators(estimator, check):
     if check.func.__name__ in SKIP_CHECKS:
-        pytest.skip("Ordcomp ordinal embedding estimator's are not fully compatible to sklearn estimators.")
+        pytest.skip("cblearn ordinal embedding estimator's are not fully compatible to sklearn estimators.")
 
     with wrap_triplet_estimator(estimator) as wrapped_estimator:
         check(wrapped_estimator)

ordcomp/embedding/wrapper/_mlds.py → cblearn/embedding/wrapper/_mlds.py

@@ -4,9 +4,9 @@
 from sklearn.utils import check_random_state
 import numpy as np
 
-from ordcomp import utils
-from ordcomp.embedding._base import TripletEmbeddingMixin
-from ordcomp.embedding.wrapper._r_base import RWrapperMixin
+from cblearn import utils
+from cblearn.embedding._base import TripletEmbeddingMixin
+from cblearn.embedding.wrapper._r_base import RWrapperMixin
 
 
 class MLDS(BaseEstimator, TripletEmbeddingMixin, RWrapperMixin):
@@ -24,7 +24,7 @@ class MLDS(BaseEstimator, TripletEmbeddingMixin, RWrapperMixin):
         embedding_: array-likeThe final embedding, shape (n_objects, 1)
         log_likelihood_: The final log-likelihood of the embedding.
 
-    >>> from ordcomp import datasets
+    >>> from cblearn import datasets
     >>> triplets = datasets.make_random_triplets(np.arange(15).reshape(-1, 1), size=400, result_format='list-order')
     >>> triplets.shape, np.unique(triplets).shape
     ((400, 3), (15,))

ordcomp/embedding/wrapper/_r_base.py → cblearn/embedding/wrapper/_r_base.py

@@ -17,8 +17,8 @@ def init_r(cls):
             numpy2ri.activate()
         except ImportError:
             raise ImportError("Expects installed python package 'rpy2', could not find it. "
-                              "Did you install ordcomp with the r_wrapper extras? "
-                              "pip install ordcomp[r_wrapper]")
+                              "Did you install cblearn with the r_wrapper extras? "
+                              "pip install cblearn[r_wrapper]")
 
     @classmethod
     def import_r_package(cls, package_name, install_if_missing=True, **kwargs):

ordcomp/embedding/wrapper/_soe.py → cblearn/embedding/wrapper/_soe.py

@@ -4,10 +4,9 @@
 from sklearn.utils import check_random_state
 import numpy as np
 
-import ordcomp
-from ordcomp import utils
-from ordcomp.embedding._base import TripletEmbeddingMixin
-from ordcomp.embedding.wrapper._r_base import RWrapperMixin
+from cblearn import utils
+from cblearn.embedding._base import TripletEmbeddingMixin
+from cblearn.embedding.wrapper._r_base import RWrapperMixin
 
 
 class SOE(BaseEstimator, TripletEmbeddingMixin, RWrapperMixin):
@@ -20,7 +19,7 @@ class SOE(BaseEstimator, TripletEmbeddingMixin, RWrapperMixin):
             stress_: Final value of the SOE stress corresponding to the embedding.
 
         Examples:
-        >>> from ordcomp import datasets
+        >>> from cblearn import datasets
         >>> triplets = datasets.make_random_triplets(np.random.rand(15, 2), result_format='list-order', size=1000)
         >>> triplets.shape, np.unique(triplets).shape
         ((1000, 3), (15,))
@@ -61,7 +60,7 @@ def __init__(self, n_components=2, n_init=10, margin=.1, max_iter=1000, verbose=
         self.random_state = random_state
 
     def fit(self, X: utils.Questions, y: np.ndarray = None, init: np.ndarray = None,
-            n_objects: Optional[int] = None) -> 'ordcomp.embedding.wrapper.SOE':
+            n_objects: Optional[int] = None) -> 'SOE':
         """Computes the embedding.
 
         Args:

ordcomp/metrics/_triplets.py → cblearn/metrics/_triplets.py

@@ -4,7 +4,7 @@
 from sklearn.utils import check_array
 from sklearn import metrics
 
-from ordcomp import utils
+from cblearn import utils
 from .. import datasets
 
 

ordcomp/metrics/tests/test_procrustes.py → cblearn/metrics/tests/test_procrustes.py

@@ -1,6 +1,6 @@
 import numpy as np
 
-from ordcomp import metrics
+from cblearn import metrics
 
 
 def test_procrustes_distance():

ordcomp/metrics/tests/test_triplets.py → cblearn/metrics/tests/test_triplets.py

@@ -1,8 +1,8 @@
 import numpy as np
 import pytest
 
-from ordcomp import metrics
-from ordcomp import datasets
+from cblearn import metrics
+from cblearn import datasets
 
 
 class DummyOrdinalEmbedding():

ordcomp/utils/_torch.py → cblearn/utils/_torch.py

@@ -16,10 +16,10 @@ def assert_torch_is_available() -> None:
         import torch  # noqa: F401  We do not use torch here on purpose
     except ModuleNotFoundError as e:
         raise ModuleNotFoundError(f"{e}.\n\n"
-                                  "This function of ordcomp requires the installation of the pytorch package.\n"
+                                  "This function of cblearn requires the installation of the pytorch package.\n"
                                   "To install pytorch, visit https://pytorch.org/get-started/locally/\n"
                                   "or run either of the following commands:\n"
-                                  "    pip install ordcomp[torch]\n"
+                                  "    pip install cblearn[torch]\n"
                                   "    pip install torch\n"
                                   "    conda install -c=conda-forge pytorch\n")
 

ordcomp/utils/tests/test_validate_data.py → cblearn/utils/tests/test_validate_data.py

@@ -2,7 +2,7 @@
 import numpy as np
 import sparse
 
-from ordcomp import utils
+from cblearn import utils
 
 
 triplets_numeric_undecided = [[0, 1, 2],

cblearn/utils/tests/test_validate_size.py

@@ -0,0 +1,14 @@
+import pytest
+
+import cblearn.utils
+
+
+def test_check_size():
+    assert 6 == cblearn.utils._validate_size.check_size(None, 6)
+    assert 3 == cblearn.utils._validate_size.check_size(3, 6)
+    assert 3 == cblearn.utils._validate_size.check_size(3., 6)
+    assert 3 == cblearn.utils._validate_size.check_size(.5, 6)
+    assert 15 == cblearn.utils._validate_size.check_size(15, 6)
+
+    with pytest.raises(ValueError):
+        cblearn.utils._validate_size.check_size(-1, 6)